Teacher–student interpersonal relationships in Indonesia: profiles and importance to student motivation

British Journal of Educational Psychology (2013), 84, 459–482

© 2013 The British Psychological Society

www.wileyonlinelibrary.com

Teacher–student interpersonal relationships dochange and affect academic motivation: Amultilevel growth curve modelling

Ridwan Maulana1,2*, Marie-Christine Opdenakker2 andRoel Bosker2

1Department of Teacher Education, University of Groningen, The Netherlands2Groningen Institute for Educational Sciences (GION), University of Groningen,

The Netherlands

Background. Research has shown that the teacher–student interpersonal relationship(TSIR) is important for student motivation. Although TSIR has received a growing

interest, there are only few studies that focus on changes and links between TSIR and

student academic motivation in a longitudinal fashion in non-Western contexts.

Aims. This study investigated changes in TSIR and links with academic motivation as

perceived by first-grade secondary school students in Indonesia. TSIR was studied from

the perspective of interpersonal behaviour in terms of Influence and Proximity. Students’

academic motivation was studied from the perspective of self-determination theory.

Sample and methods. A total of 504 first-grade secondary school students of 16

mathematics and English classes participated in the study. Surveys were administered in

five waves throughout the school year. Multilevel growth curve modelling was applied.

Results. Contrary to the (limited) general research findings fromWestern contexts, we

found that the quality of TSIR (student perceptions) increased over time. The increase

was slightly more pronounced for Proximity than for Influence. In accordance with the

findings for theWestern countries, the level of students’ controlledmotivation increased,

while that of autonomous motivation decreased over time. However, the negative

change in autonomous motivation was less pronounced. As in Western countries, TSIR

was longitudinally linked with academic motivation, in particular, with autonomous

motivation.

Conclusions. Evidence is found that TSIR can change in a favourable way, and this

positively affects student motivation. Future research could benefit from unravelling the

influences of cultures on changes in TSIR in broader contexts.

Existing studies have demonstrated that good teacher–student relationships can result in aproductive classroom environment (Davis, 2003; den Brok, Brekelmans, & Wubbels,

2004; Henderson, Fisher, & Fraser, 2000; Opdenakker, Maulana, & den Brok, 2012).

Particularly, a high-quality teacher–student interpersonal relationships (TSIR) enhancesstudent motivation; conversely, a poor TSIR is often reflected in a decline in students’

*Correspondence should be addressed to Ridwan Maulana, Department of Teacher Education, Landleven 1, 9747 ADGroningen, The Netherlands (emails: [email protected]; [email protected]).

DOI:10.1111/bjep.12031

459

motivation for learning (den Brok et al., 2004; Opdenakker et al., 2012; Wubbels &

Brekelmans, 2005).

Despite the numerous studies indicating the importance of TSIR for academic

motivation, the knowledge acquired thus far shows several limitations. First, we stilllack a thorough understanding of how contextual changes in, in particular, TSIR

influences motivation over time: Studies on the topic usually apply a limited

cross-sectional or short-term design. Second, studies that do pay attention to changes

in TSIR over time tend to focus on the Western context only. In East Asian countries,

such as Indonesia, the influence of TSIR and changes in TSIR has hardly ever been

researched. One cross-sectional study of the Indonesian situation suggests that TSIR is

an important predictor of student motivation in this setting as well (Maulana,

Opdenakker, den Brok, & Bosker, 2011), yet it is unclear whether TSIR shows similartrend patterns and has a similar effect on student motivation, as in Western countries.

As a third shortcoming, there are only few studies that focus on the first year of

secondary schooling (Exceptions are: Maulana, Helms-Lorenz, & Van de Grift, 2013;

Maulana, Opdenakker, den Brok, & Bosker, 2012a; Maulana, Opdenakker, Stroet, &

Bosker, 2013; Opdenakker & Maulana, 2010). This particular year is of prime

importance to students, however, because it forms the transition from primary to

secondary school, constituting a phase of new experiences and, also, the period when

the TSIR is first established and developed. Indeed, the first year of secondary schoolcan be said to offer a unique time frame for studying the dynamics between TSIR and

academic motivation.

This study aims to add to existing knowledge by (1) investigating students’ perception

of TSIR and their motivation as they develop during the first year of secondary school and,

also, examining how these two factors are longitudinally linked; (2) exploring the effects

of several personal and contextual characteristics on differences and changes in TSIR and

academic motivation over time; (3) gathering data on TSIR in a non-Western context

(i.e., Indonesia); and (4) applying multilevel growth curve models to handle thehierarchical structure of the data, whilst paying attention to the variance, differences, and

changes in TSIR and academic motivation over time.

Changes in TSIR over time

The literature on TSIR is extensive, showing a wide variety of measures and approaches

being applied to quantify and qualify the nature of the relationship. Likewise, the

importance of TSIR has been studied inmany different fields of research, such as, researchon learning environments, teaching and teacher education, child development and social

development, motivational theory, and teacher/classroom effectiveness (den Brok,

VanTartwijk, Wubbels, & Veldman, 2010; Jerome, Hamre, & Pianta, 2008; Maulana,

Helms-Lorenz, et al., 2013; Maulana, Opdenakker, et al., 2013; Opdenakker et al., 2012).

Usually, the concept of TSIR is studied by focusing on its various components, which can

include teacher–student closeness (Davis, 2003; Newberry & Davis, 2008), participation

in the classroom (Davis, 2003; Turner & Patrick, 2004), classroom control, fairness

(Wentzel, 2002), classroom belonging (Anderman, Patrick, & Ryan, 2004; Freeman,Anderman, & Jensen, 2007), and teacher involvement (Maulana, Helms-Lorenz, et al.,

2013; Maulana, Opdenakker, et al., 2013). In previously conducted studies, all TSIR

elements mentioned here have been shown to be positively related to academic

motivation, students’ attitude towards certain subjects, and their academic achievements

(den Brok et al., 2010; Opdenakker et al., 2012).

460 Ridwan Maulana et al.

In the present study, students’ perception of TSIR is studied from the perspective of

interpersonal relationships such as specified by the Model of Interpersonal Teacher

Behaviour (MITB: Wubbels & Brekelmans, 2005; see Figure 1), a theoretical framework

which is based on the circumplex model of Leary (1957). According to the MITB,interpersonal behaviour canbemapped in termsof twodimensions, namely Influence and

Proximity. Influence refers to behaviours associated with teacher control and authority

(with the opposing values of Dominance–Submission). Proximity refers to behaviours

indicative of the closeness of interpersonal ties, such as teacher friendliness, how warm

the teacher–student relationship is, and the care shown by teachers (Cooperation–Opposition). By applying these two dimensions, each of which can assume two values

(Dominance vs. Submission and Cooperation vs. Opposition), a coordinate system can be

formed of eight cross-sections describing different types of interpersonal teacherbehaviour. The different types distinguished here are (1) Leadership (DC), (2) Helpful/

Friendly (CD), (3) Understanding (CS), (4) Student Freedom (SC), (5) Uncertain (SO), (6)

Dissatisfied (OS), (7) Admonishing (OD), and (8) Strict (DO; see also Figure 1).

The two dimensions of interpersonal behaviour specified in the MITB have been

recognized as primary influences on the social interaction in an educational context (den

Brok et al., 2010; Wubbels, Brekelmans, den Brok, & van Tartwijk, 2006). Conceptually,

these dimensions encompass the TSIR components used in previous research, such as

closeness, support, control, belonging, and participation (den Brok et al., 2010). Inaddition, there is evidence that theMITB can easily be applied in different cultural settings

(Maulana, Opdenakker, den Brok, & Bosker, 2012b; Triandis, 1993).

To measure students’ perception of teachers’ interpersonal behaviour based on

the MITB framework, a questionnaire on teacher interaction (QTI) was constructed

Figure 1. Model of interpersonal teacher behaviour (Wubbels & Brekelmans, 2005).

Interpersonal relationship and academic motivation 461

(for a review of studies using the samemeasuring instrument, seeWubbels & Levy, 1993;

Wubbels et al., 2006). TheQTI has been validated in different cultural contexts, including

the Netherlands, the USA, Australia, Singapore, China, Korea, and Indonesia (for an

overview, see Maulana et al., 2012a,b). In the research domain of teacher effectivenessand learning environments, students’ perceptions of the two dimensions of teacher

interpersonal behaviour have shown to be consistently related to student motivation and

academic achievements (Opdenakker et al., 2012; Wubbels et al., 2006).

A limited number of studies have documented howTSIR develops in both primary and

secondary education. Some studies suggest relatively stable trends in TSIR development

over time (Brekelmans, 1989; Curby, Grimm, & Pianta, 2010; Jerome et al., 2008;

Mainhard, Brekelmans, den Brok, & Wubbels, 2011; Patrick, Turner, Meyer, & Midgley,

2003; Ryan & Patrick, 2001; Skinner & Belmont, 1993). However, unstable tendencieshave been recorded as well, with the quality of TSIR deteriorating as time progresses

(Evertson & Veldman, 1981; Flanders, Morrison, & Brode, 1968; Flores, 2003; Huang &

Waxman, 1992; Maulana, Helms-Lorenz, et al., 2013; Maulana, Opdenakker, et al., 2013;

Opdenakker&Maulana, 2010;Opdenakker et al., 2012; Pianta&Stuhlman, 2004; Skinner

& Belmont, 1993). Research has shown that the sooner teachers establish a positive

interpersonal relationship with their students, the more likely it is that the decline in the

quality of TSIR is prevented (Opdenakker et al., 2012; Patrick et al., 2003; Ryan&Patrick,

2001).Usually, studies onTSIR are conducted inWestern countries.Moreover, research tends

to focus on relatively short time periods (up to 4 months or shorter) or, in the case of

longer periods (6 months to a year), it involves only two measuring moments. The

common approach is to study students’ perceptions, teachers’ perceptions, or a

combination of these two approaches. Interesting enough, stability or deterioration of

TSIR can be observed in both short- and long-term studies. How exactly students’

perception of TSIR develops over time remains unclear, however. To our knowledge, no

research has been carried out as yet which includes multiple measurements taken in thecourse of a given school year. Also, information on TSIR development in non-Western

contexts is still lacking.

Our study will help to fill these three gaps in the existing knowledge on TSIR,

examining howTSIR changes over time by takingmultiplemeasurements, relating TSIR to

academic motivation, and exploring these patterns in a non-Western context. As such, it

makes a clear andmuch-needed contribution to the study of changes in TSIR and its effects

on academicmotivation and stresses the universal applicability and the dynamic nature of

TSIR and academic motivation.

Changes in academic motivation over time

The conceptual framework we apply for studying student motivation reflects the

theoretical idea of there being a motivational continuum, a notion which derives from

Self-Determination Theory (SDT; Ryan & Deci, 2000). SDT recognizes four types of

regulation styles, indicative of different degrees of self-determination. These four types are

external, introjected, identified, and integrated regulation. Eachof these regulation stylesis theoretically conceptualized as a form of extrinsic motivation.

External regulation is the least autonomous regulation style, having an external

perceived locus of causality. Introjected regulation is relatively controlled: Even though

regulation is internally driven, the perceived locus of causality is external and not

completely experienced as originating from the self. Identified regulation is more

462 Ridwan Maulana et al.

autonomous and self-determined: Identification reflects a consciousness of the value of

regulation, and interaction is seen as personally important. Finally, integrated regulation

is the most autonomous form of motivation, with regulation being fully assimilated with

the self. Although integrated regulation is theoretically distinguishable from intrinsicmotivation – it is instrumental and therefore extrinsic – integrated regulation shares

many of the qualities of intrinsic motivation, being both autonomous and ‘unconflict-

ing’ (Ryan & Deci, 2000). In empirical research, integrated regulation is often

operationalized in the same manner as intrinsic motivation (Ryan & Connell, 1989;

Vansteenkiste, Simons, Lens, Sheldon, & Deci, 2004; Vansteenkiste, Zhou, Lens, &

Soenens, 2005).

Motivation is generally recognized as playing an important role in students’ learning

and academic achievement, yet research has shown that motivation tends to wane asstudents grow older andmove to higher grades. The transition from primary to secondary

education is thought to be an especially critical phase (Anderman, Maehr, & Midgley,

1999; Maulana et al., 2012a; Murphy & Alexaneder, 2000; Yeung, Lau, & Nie, 2011). In

particular intrinsicmotivation is likely to decrease (Corpus,McClintic-Gilbert, &Hayenga,

2009; Opdenakker et al., 2012; Otis, Grouzet, & Pelletier, 2005), declining ever more as

time progresses (Gottfried, Fleming, & Gottfried, 2001; Lepper, Corpus, & Iyengar, 2005;

Spinath & Spinath, 2005). The findings on trends in extrinsic motivation are somewhat

inconclusive. Some studies have established that extrinsic motivation declines during theprocess of schooling (Lepper et al., 2005; Otis et al., 2005; Ratelle, Guay, Larose, &

Senecal, 2004), whereas other studies suggest that it increases (Anderman et al., 1999;

Harter, 1981).

Notably, most studies referred to above base themselves on cross-sectional designs, or

onmeasurements taken in different school years (between-year changes). Although there

are some studies that do take a more longitudinal approach and focus on within-year

changes, measurement is often limited to just two waves within a limited time frame. An

accurate description of developmental paths over the course of an entire school yeartherefore cannot be given.

Linking changes in TSIR and academic motivation over time

The development of students’ motivation results partly from their perception of the social

climate in the classroom. Research has demonstrated that there is a clear link between

students’ attitude and teachers’ encouragement (Flanders et al., 1968) and that an

unfavourable social climate in the classroom can lead to a decline in students’ motivation(Feldlaufer, Midgley, & Eccles, 1988), disruptive behaviour, and a loss of mutual respect

and support (Ryan & Patrick, 2001).

In order to understand the differences and longitudinal changes in students’

motivation, it is vital to recognize the role played in this by the social environment of

the classroom (Opdenakker & Maulana, 2010). Teachers who display encouraging

behaviour can have a positive effect on students’ engagement and motivation

(Corpus et al., 2009; Skinner & Belmont, 1993). A supportive TSIR can even partly

counter the often-observed decline in students’ autonomous motivation as timeprogresses (Lapointe, Legault, & Batiste, 2005; Wentzel, 2010). Stated differently, a

high-quality TSIR can serve as a form of protection (Opdenakker et al., 2012). All in

all, therefore, the research on TSIR suggests that a good social climate in the

classroom makes it more likely that students’ interest and learning ability develop in

a positive manner.

Interpersonal relationship and academic motivation 463

Contextual and personal characteristics

There are a number of contextual and personal characteristics that can act as predictors of

differences and changes in TSIR and student motivation. For instance, the learning

environment is often better in high ability compared with other types of classes (Boufard&Couture, 2003; Evertson, 1982; Lapointe et al., 2005;Maulana et al., 2012a;Mills, 1997;

Opdenakker & Van Damme, 2001). At the same time, evidence has been found as well to

suggest that teachers’ cooperativeness and students’ autonomous motivation decline

faster in high-ability classes than in classes of mixed ability and, also, that controlled

motivation only increases in high-ability classes (Opdenakker et al., 2012). In addition,

science andmaths teachers are often judged in less favourable terms than teachers of other

subjects (Levy, den Brok, Wubbels, & Brekelmans, 2003; Maulana et al., 2012b).

Studies on several other aspects of the classroom climate have shown that, in general,girls take a more positive view of teachers than boys (Fraser, 2007). Girls also perceive

their teachers more often as dominant (teacher-centred) and cooperative personalities

(Levy et al., 2003; Wubbels & Levy, 1993). Boys stereotypically more often experience a

sense of being in conflict with their teachers, whilst girls tend to report greater

interpersonal closeness (Hamre & Pianta, 2001; Ladd, Birch, & Buhs, 1999; Saft & Pianta,

2001). Notably, Levy et al. (2003) also found that boys see male teachers as more helpful,

friendly, and understanding than female teachers; however, their study gives no

indication that girls see female teachers as more helpful and understanding.1

As another factor to be mentioned, early closeness with teachers reduces conflict

between students and teachers at a later stage. Having said as much, a high teacher

dependency in the initial stages often leads to more conflict between boys and teachers

later on (Hamre & Pianta, 2001). With regard to gender effects on motivation, girls are

usually more highly motivated than boys (Opdenakker et al., 2012).2 Gender differences

can be discerned aswell in the trends or development of studentmotivation, although the

literature is rather inconclusive on this point. Some studies find no difference between

girls and boys (Meece &Miller, 2001), whereas others describe steeper downward trendsin motivation for either boys (Van de Gaer et al., 2007; Van der Werf, Opdenakker, &

Kuyper, 2008) or girls (Opdenakker et al., 2012).

The Indonesian context: General pictures

Contrary to the individualist culture of most Western countries, Indonesian society is

collectivist in nature (Hofstede, 1991). In a collectivist society, interpersonal closeness, as

expressed by close and regular contact between individuals in everyday life, is a highlyvalued commodity. The classroom context in particular reflects such collectivist

tendencies, constituting, in effect, a social unit within the larger unit of society.

Moreover, in the Indonesian context, TSIR is influenced by other pervasive cultural

values, such as the idea of paternalism and respect for the authority of one’s elders, which

implicitly regulate the interaction between the young and the old (Liem, Martin, Nair,

1 The study conducted by Levy et al. (2003) suggests that the relation between gender and the perception of teacher behaviour isnot straightforward: both factors interact with a number of other variables, such as curricular and subject preference (Baker &Leary, 1995; Kahle, Parker, Rennie, & Riley, 1993), and culturally related gender role definitions (Marcus, Gross, & Seefeldt,1991; Timm, 1999).2 Some studies indicate that gender differences in motivation may be due to role stereotypes. Although motivation generallydeclines during adolescence, gender differences, such as boys favouring traditionally male-dominated subjects (e.g., maths andscience) and girls favouring so-called female subjects (e.g., language), are often found (Ireson & Hallam, 2009; Maulana et al.,2011; Midgley, Kaplan, & Middleton, 2001; Usher & Pajares, 2009; Watt, 2008).

464 Ridwan Maulana et al.

Bernardo, & Prasetya, 2009; Maulana et al., 2011). As a consequence, the perceived

distancebetween student and teacher is relatively high; also, the school system tends to be

hierarchical and allows for little variation.

Less than a decade ago, the Indonesian government implemented a number of changesin the national schooling system, moving from a centralized to a more decentralized, and

less uniform, curriculum (Mullis et al., 2008). Shortly after these changes were made, the

government also introduced a new scheme for categorizing secondary schools based

upon their standard of qualification. Schools are divided into (1) schools of international

standard (highest qualification), (2) pilot schools of international standard (second

highest qualification), and (3) schools of national standard (lowest qualification). The

distinction between homogeneous classes (either high-ability or low-ability pupils only)3

and heterogeneous classes (mixed-ability) follows mainly from a school’s standard ofqualification. Some schools offer both homogeneous and heterogeneous classes, but

there are also schools with just homogeneous or heterogeneous classes (Ministry of

National Education, 2007).

Aim and research questions

Based on previous research on TSIR and academic motivation reviewed earlier, our study

aims to (1) investigate changes in TSIR during the first year of secondary education in

Indonesia and (2) examine the longitudinal links such as exist betweenTSIR and academic

motivation. To be precise, we will be addressing the following research questions:

(1) How variable is TSIR, between classes and over measurement occasions?

(2) Does TSIR change across the school year?

(3) Are there differences between classes as regards change over time?

(4) How do personal and contextual characteristics affect TSIR over time?

(5) How is TSIR linked with students’ academic motivation over time?

In the discussion, we will discuss findings of this study also in relation with findings

from Western contexts. This will give the opportunity to make a comparison with

different contexts. However, it is not our intention to do a cross-cultural study. Weacknowledge instead that there are important context differences between Indonesia and

Western countries.

Method

Participants

A total of 504 first-grade students (56% girls; all aged 11–13) from eight mathematics and

eight English as a Foreign Language (EFL) classes (52% heterogeneous) from four

secondary schools in West-Indonesia participated in the study. Classes within schools

were randomly selected with a convenience sampling procedure, resulting in a group of

10 teachers, 80% of which was female. Of the 10 teachers selected, six were maths

teachers and four were EFL teachers. One of the maths teachers was male, and five were

3 Theoretically speaking, homogeneous school classes refer to either high-ability groups (corresponding to international oracceleration classes) or low-ability groups (corresponding to regular classes). In our sample, however, we confined ourselves tohomogeneous classes of high-ability students and international classes of second qualification schools. Usually, high-abilitystudents will have chosen a school of either the first or second standard of qualification. In practice, though, high-ability studentscan also choose to go to a lower standard school, because, for instance, it is cheaper. For this reason, the third category of schoolssometimes has mixed-ability classes.

Interpersonal relationship and academic motivation 465

female. Of the EFL teachers, one was male and three were female. The teachers’

professional experience ranged from one to over 20 years.

Measures

Teacher–student interpersonal relationshipWe used the Indonesian version of the QTI (Maulana et al., 2011, 2012b) to measurestudents’ perception of teacher Influence and Proximity. The measure is based on the

MITB framework described earlier (Wubbels et al., 2006). The questionnaire was

originally formulated in Dutch and subsequently translated into English (Wubbels & Levy,

1993). The Indonesian version was developed in accordance with the guidelines of the

International Test Commission (Hambleton, 1994). In order to meet the specific

requirements of our study, we adapted the questionnaire in such a way that it could be

used tomeasure students’ perception of Influence and Proximity inmaths and EFL classes.

The ultimate questionnaire consisted of 56 items measuring eight different interpersonalstyles,with answer categories ranging from (1)never to (5)always. Examples of the items

used are ‘This teacher is friendly’, ‘This teacher is hesitant’, or ‘This teacher is strict’.

The reliability of theQTI scaleswas satisfactory acrossmeasurements, ranging from .65

to .72 at the student level to .70–.94 at class level. Intraclass correlation scores varied from.19 to .27, confirming that the scales used can distinguish sufficiently between classes.

A confirmatory factor analysis (with Mplus; Muthen & Muthen, 1999) generated an ideal

circumplex model with a reasonable fit (v2 = 181.59 with p = .00; CFI = .99; TLI = .99;

RMSEA = .06 and SRMR = .08), although there does remain some room for improve-ment.4 In general, the results suggest that the psychometric quality of the measures we

used is good and that they concord with the quality of the original Dutch and English

versions. The set of general descriptive statistics calculated shows that the teachers scored

highly moderate on Influence and slightly positive on Proximity (see Table 1).

Academic motivation

We measured academic motivation by means of the questionnaire for motivationaldimensions (Vansteenkiste et al., 2004). This measure is based on the academic

self-regulation scale of Ryan and Connell (1989) and assesses students’ reasons for

studying a particular subject. The questionnaire used consisted of 16 items, with four

items per scale. The scales applied were, respectively: (1) external regulation (with items

such as ‘I study this subject because my parents ask me’), (2) introjected regulation

4 In the theoretical circumplex model of the MITB, the eight scales are expected to be ordered in a circular structure. The eightscales are cross-sectionally derived from the two dimensions (or factors): Influence and Proximity. In such a model, the factorloadings represent coordinates in the circular structure. Accordingly, each scale (or item) is expected to load on both factors at thesame time, even though they can vary in magnitude. This aspect of the MITB diverges from regular factor models, the latterin which scales (or items) are expected to load (positively or negatively) on one factor only. The theoretical dimensionscores on Influence and Proximity were computed as follows: Influence = (.92*DC) + (.38*CD) � (.38*CS) �(.92*SC) � (.92*SO) � (.38*OS) + (.38*OD) + (.92*DO); Proximity = (.38*DC) + (.92*CD) + (.92*CS) + (.38*SC) �(.38*SO) � (.92*OS) � (.92*OD) � (.38*DO). In these equations, the coefficients are the expected factor loadings; theabbreviations represent the scale scores. Dimension scores can vary between�3 and+3. Themore positive the dimension score,the more positively students perceive a teacher’s interpersonal behaviour to be (den Brok, Brekelmans, & Wubbels, 2006; denBrok et al., 2010). With regard to Influence, a score of�3 suggests highly submissive behaviour (very little Influence), whereas+3 is indicative of highly dominant behaviour (very high Influence). As to Proximity,�3 reflects highly hostile or opposing behaviour(very low Proximity) and +3 reflects highly cooperative behaviour (very high Proximity). Generally speaking, previous empiricalstudies have shown that, in practice, the dimension scores tend to lie between 0 and 1 (Maulana et al., 2012b).

466 Ridwan Maulana et al.

(e.g., ‘I study this subject because I would feel ashamed if I don’t do it’), (3) identified

regulation (e.g., ‘I study this subject because it is personally important to me’), and (4)

intrinsic motivation (e.g., ‘I study this subject because I find it interesting’). The first two

scales refer to a more controlled regulation style of motivation (controlled motivation);

the other two scales are indicative of a more autonomous regulation style of motivation

(autonomous motivation). We used 5-point Likert scales with answers ranging from 1

(completely false) to 5 (completely true).

The reliability of our measures was satisfactory across measurements, with thefollowing a-values: (1) external regulation: a = .62–.69, (2) introjected regulation:

a = .62–.68, (3) identified regulation: a = .63–.82, and (4) intrinsic motivation:

a = .81–.87. Following the studybyVansteenkiste et al. (2004),weconstructed composite

scores for controlled motivation (a = .68–.72) by averaging the external and introjected

regulation scales; in a likewise manner, we constructed values for autonomous motivation

(a = .82–.90) by averaging the identified regulation and intrinsic motivation scales.5

Overall, the descriptive statistics show that students’ controlled and autonomous

motivations are highly moderate, with the highest scores on autonomous motivation(see Table 2).

Contextual and personal characteristics

We coded time in accordance with the moments of measurement in the survey (in

months): 0 (baseline), 1.5, 4, 7, and 10. We divided the classes of students into two

categories, with ‘0’ referring to a high-ability class (homogeneous) and ‘1’ referring to a

mixed-ability class (heterogeneous). The variables teaching subject and student genderwere included in the analyses as dummy variables (with the values of, respectively, ‘0’ for

mathematics and ‘1’ for EFL, and ‘0’ for boys and ‘1’ for girls).

Procedure

Prior to the survey, we gave the sampled students comprehensive instructions on how to

use a 5-point Likert scale for responding to the questionnaire items. The actual surveys

Table 1. Descriptive statistics of teacher Influence and Proximity across measurementsa

Dimension M SD Min Max

Influence 0.42 0.36 �0.73 1.55

Proximity 0.73 0.65 �1.77 2.43

Note. an = 504. Dimension scores of Influence and Proximity range between�3 and +3. These numbers

represent, in case of Influence, the degree of Influence from �3 (submission) to +3 (dominance) and, in

case of Proximity, the degree of Proximity from�3 (opposition) to+3 (cooperation). The range of scoresis 0–0.5 (moderately positive), 0.5–1 (positive), and >1 (very positive; see den Brok et al., 2006 for a

detailed overview of the dimension scores).

5 Both identified regulation and intrinsic motivation are characterized by the willingness to engage in a certain activity, eitherbecause someone perceives the contents of a subject as fitting their own values (identified regulation) or because they find itinteresting in itself (intrinsic regulation). Both forms ofmotivation share the characteristic of inspiring an unconflicted and volitionalform of functioning. This justifies using a combination of the two types of motivation, even though identified regulation is anextrinsic type of motivation. Autonomous motivation can subsequently be contrasted with more instigated or controlled form ofmotivation (Vansteenkiste et al., 2004, 2005).

Interpersonal relationship and academic motivation 467

were conducted during the school year of 2009–2010. The first survey, carried out in

July,6 aimed to capture students’ initial motivation, so at themoment of starting secondary

education (baseline). Subsequent waves were spread evenly over the rest of the school

year (September, December, March, and June). The teachers were not present whenstudents filled in their questionnaires.

Analytical approach

We applied multilevel growth curve modelling (MGCM, with MLwiN; Rasbash, Steele,

Browne, & Goldstein, 2009) to analyse the trends in TSIR and academic motivation. The

models we used comprised three levels, namely the measurement occasion (level 1), the

student (level 2), and class (level 3).Modelling took place in a stepwise approach: We began by estimating empty models,

after which we added the time component (linear to polynomial terms) so to arrive at an

adequate estimation of the actual data. We studied the differences in changes between

classes and between students by introducing the random effect of time. In a subsequent

phase, personal, contextual, and TSIR predictors were added, to model the different

motivational constructs. The fixed effects in the model were tested using t-ratio

coefficients and by applying the rule that an absolute t-value should be greater than�1.64

(p < .10) for a variable to have a significant effect (Snijders & Bosker, 1999). This analyticstrategy was applied for each of the separate measures.

Results

Variability of TSIR

For Influence, we found that 3% of all variance could be attributed to class level, 12% tostudent level, and 85% tomeasurement occasion level. For Proximity, we found that 7% of

all variance could be attributed to class level, 23% to student level, and 70% to

measurement occasion level. As this indicates, only a moderate difference can be said to

exist between classes with respect to Influence and Proximity. Furthermore, both

dimensions are not stable over time. Comparedwith Proximity, Influence seems to be the

less stable dimension.

Changes in TSIR over time

The MGCM suggests that the best way of representing the changes in Influence and

Proximity is a positive linear trend (see Table 3, Figure 2). Both Influence and

Proximity increase over time (bs = .01 and .02; p < .05).

Table 2. Descriptive statistics of controlled and autonomous motivation across measurementsa

M SD Min Max

Controlled 3.55 0.70 1.00 5.00

Autonomous 4.22 0.69 1.00 5.00

Note. an = 504.

6 In Indonesia, the school year starts in July and ends in June, with a break of a few weeks between successive school years.

468 Ridwan Maulana et al.

Class differences in change over timeRegarding the differences between classes, the results show that the 95% interval includes

both negative and positive time (linear) effects. Recalculating the interval limits for a

period of 10 months (which corresponds to a regular school year), the 95% interval ranges

between �0.31 and 0.59 for Influence and between �0.52 and 0.92 for Proximity.

Table 3. Results of multilevel models of the development of teacher Influence and Proximity over time

with time random at class and student level

Variable

Teacher Influencea Teacher Proximityb

Coefficient SE Coefficient SE

Fixed effect

Intercept 0.3393*** 0.0306 0.6304*** 0.0727

Time 0.0138* 0.0062 0.0197* 0.0099

Time2

Random effect

Level 3 variance (class)

Intercept 0.0112 0.0054 0.0734 0.0303

Intercept 9 time �0.0022 0.0010 �0.0074 0.0037

Time 0.0005 0.0002 0.0013 0.0006

Level 2 variance (student)

Intercept 0.0000 0.0000 0.1070 0.0116

Intercept 9 time 0.0000 0.0000

Time 0.0000 0.0000

Level 1 variance (occasion)

Residual 0.1213 0.0039 0.3836 0.0122

Deviance 1515.02 3661.09

Difference in deviance

(compared to model with

time as a fixed effect)

17.35; d.f. = 4; p < .001 65.27; d.f. = 4; p < .001

Note. an = 2,008. bn = 2,009.

°p < .10. *p < .05; **p < .01; ***p < .001.

Figure 2. General trajectories of Influence and Proximity during the first year of secondary education

(raw scores).

Interpersonal relationship and academic motivation 469

This indicates substantial between-class differences on both dimensions. The differences

in Proximity seem to be larger than the differences in Influence.

Personal and contextual characteristics

Differences in TSIR can partly result from the influence of student gender and class type

(see Table 4). With regard to Influence, the main effect of student gender is positively

significant (b = .10, p < .01), suggesting that girls report a higher level of Influence thanboys. Moreover, the effect of the interaction between time and class type on Influence is

significant and positive (b = .03, p < .001),7 showing that in mixed-ability classes,

Influence increases faster than in high-ability classes (see Figure 3). In fact, the level of

Influence is almost constant throughout the school year (with only a small decline) in

high-ability classes; in mixed-ability classes, on the contrary, it increases as time

progresses. For Proximity, the main effect of class type was found to be negatively

significant (b = �.30, p < .01). This means that students in mixed-ability classes report

less Proximity than students in high-ability classes.

Table 4. Results of final multilevel models of the development of teacher Influence and Proximity over

time with effects of predictor variables

Variable

Teacher Influencea Teacher Proximityb

Coefficient SE Coefficient SE

Fixed effect

Intercept 0.3362*** 0.0538 0.8015*** 0.1083

Time �0.0041 0.0080 0.0258 0.0179

Time2

Subject taught �0.0317 0.0577 �0.1403 0.1178

Time 9 subject taught 0.0062 0.0096 �0.0053 0.0199

Class type �0.0721 0.0577 �0.3046** 0.1179

Time 9 class type 0.0330*** 0.0096 0.0017 0.0199

Student gender 0.0966** 0.0317 0.0833 0.0565

Time 9 student gender �0.0028 0.0049 �0.0065 0.0074

Random effect

Level 3 variance (class)

Intercept 0.0095 0.0047 0.0438 0.0198

Intercept 9 time �0.0016 0.0008 �0.0080 0.0033

Time 0.0003 0.0001 0.0014 0.0006

Level 2 variance (student)

Intercept 0.0000 0.0000 0.1067 0.0116

Intercept 9 time 0.0000 0.0000

Time 0.0000 0.0000

Level 1 variance (occasion)

Residual 0.1197 0.0038 0.2782 0.0102

Deviance 1469.85 3622.69

Note. an = 2,008. bn = 2,009.

°p < .10. *p < .05; **p < .01; ***p < .001.

7 The main effect of time (linear) was not significant any longer when the interaction effect between time and class type wasincluded in the model of Influence and Proximity. This suggests that the effect of time is associated with the effect of class type.

470 Ridwan Maulana et al.

Longitudinal links between TSIR and academic motivation

With regard to controlled motivation, we found that 7% of all variance could be attributed

to class level, 26% to student level, and 67% to measurement occasion level. For

autonomous motivation, these percentages were 5%, 23%, and 72%. As it turns out, then,

student motivation is not stable over time; also, autonomous motivation is the less stable

component of student motivation.

TheMGCM results show that controlledmotivation increases as the school year draws

on, with a small deceleration as time progresses, as shown by the linear and quadraticeffects of time (see Table 5, Figure 4). A systematic increase can be seen during the first

7 months; controlled motivation decreases slightly after this period. Inspection of

between-class differences related to the overall linear trend reveals that the 95%

confidence interval includes only positive effects of time. Estimating the interval limits for

period of 10 months, the 95% interval varies from 0.26 to 0.66.

Student gender and the subject taught8 were shown to have no effect on controlled

motivation, contrary to class type (see Table 5, Figure 5). With respect to the latter,

students in mixed-ability classes report higher controlled motivation than students inhigh-ability classes (b = .34, p < .001). Differences in controlled motivation are linked

with differences in TSIR (see Table 5, Models 2 and 3): Both Influence and Proximity have

a significant effect on controlled motivation (bs = .15 and .10, p < .05 and .01

respectively). This suggests that a higher level of Influence and Proximity corresponds

with a higher level of controlledmotivation.9 In addition, our results show that the effects

of Influence and Proximity on controlled motivation are much more pronounced in

high-ability classes than in mixed-ability classes. In fact, Influence and Proximity are far

more closely associated with controlled motivation in high-ability classes (see Table 5,Model 4).10 There seems to be a common effect of Influence and Proximity on controlled

motivation.

Figure 3. The developmental trajectories of teacher Influence and Proximity according to class type

based on best-fitted multilevel models.

8We did test possible interaction effects between student gender and subject on academic motivation, but the results were notsignificant.9 Additional analyses showed that teacher Influence has a stronger effect on introjected regulation than on external regulation.This suggests that, although TSIR has a significant effect on controlled motivation, its effect is more pronounced for a moreself-determined (introjected) regulation than externally determined (external) regulation.10 bInfluence for heterogeneous = .1493�.1514 = �.0021, bProximity for heterogeneous = .0994�.0792 = .0202.

Interpersonal relationship and academic motivation 471

Table

5.Resultsofmultilevelmodelsofthedevelopmentofcontrolledmotivationovertime

Variable

Model1

a

Controlledmotivation

Model2

b

Controlledmotivationwith

Influence

controlvariables

Model3

c

Controlledmotivationwith

Proximityandcontrol

variables

Model4

d

Controlledmotivationwith

TSIRandcontrolvariables

Coefficient

SECoefficient

SECoefficient

SECoefficient

SE

Fixedeffect

Intercept

3.2569***

0.0626

3.2032***

0.0860

3.1782***

0.0861

3.1655***

0.0868

Time

0.0457**

0.0140

0.0547**

0.0205

0.0513*

0.0206

0.0523*

0.0207

Time2

�0.0028*

0.0011

�0.0039**

0.0015

�0.0038*

0.0015

�0.0039**

0.0015

Subject

taught

�0.0001

0.0600

0.0314

0.0711

0.0318

0.0701

0.0347

0.0696

Time9

subject

taught

�0.0008

0.0085

�0.0055

0.0106

�0.0037

0.0106

�0.0045

0.0106

Class

type

0.3379***

0.0602

0.2946***

0.0781

0.3075***

0.0778

0.3235***

0.0792

Time9

classtype

�0.0050

0.0085

0.0092

0.0107

0.0109

0.0107

0.0107

0.0107

Studentgender

0.0215

0.0519

0.0072

0.0647

0.0105

0.0647

0.0080

0.0647

Time9

studentgender

0.0036

0.0068

0.0056

0.0081

0.0056

0.0081

0.0059

0.0081

Influence

0.1493*

0.0694

0.0691

0.0809

Proximity

0.0994**

0.0358

0.0811*

0.0418

Class

type9

Influence

�0.1514°

0.0848

0.0889

0.0972

Class

type9

Proximity

�0.0792°

0.0471

�0.0553

0.0538

Random

effect

Level3

variance

(class)

Intercept

0.0039

0.0051

0.0041

0.0073

0.0034

0.0071

0.0031

0.0069

Intercept9

time

0.0005

0.0005

0.0001

0.0008

0.0002

0.0008

0.0002

0.0008

Time

0.0001

0.0001

0.0002

0.0002

0.0002

0.0002

0.0002

0.0002

Level2

variance

(student)

Intercept

0.1619

0.0219

0.1950

0.0346

0.1966

0.0347

0.1953

0.0346

Intercept9

time

�0.0054

0.0025

�0.0069

0.0040

�0.0069

0.0040

�0.0068

0.0040

Time

0.0009

0.0004

0.0009

0.0006

0.0009

0.0006

0.0009

0.0006

Level1

variance

(occasion)

Residual

0.3008

0.0115

0.2622

0.0123

0.2622

0.0123

0.2624

0.0123

Deviance

4558.02

3488.46

3484.99

3484.14

Note.

Therandomeffectsoftimeandtimequadraticareextremelysm

all(close

tozero)because

thedisplayprecisioninthemodelsisrestrictedto

fourdigitsafterthedecimalpoint.TSIR,

Teacher–studentinterpersonalrelationship.

an=2,376,bn=1,892,cn=1,892,dn=1,892.

°p<.10,*p<.05,**p

<.01,***

p<.001.

472 Ridwan Maulana et al.

As to within-year changes in autonomous motivation, the MGCM results suggest that

autonomous motivation decreases linearly over time (see Table 6). When examining the

linear effect of time on differences between classes (based on themodelwith just the time

effect), the 95% confidence interval turns out to comprise both negative and positive time

effects. The estimates of the interval limits across the school year ranged from �0.42 to

0.14, suggesting only moderate between-class differences.Class type was found to have a significant effect on autonomous motivation (see

Table 6, Figure 5). Students inmixed-ability classes report higher autonomousmotivation

than students in high-ability classes (b = .11, p < .10).Moreover, the interaction between

time and class type was shown to be significant (b = .02, p < .01), which indicates that

autonomous motivation decreases faster in high-ability classes than in mixed-ability

classes. In fact, the autonomous motivation of students in mixed-ability classes remains

fairly stable over time.

These differences in autonomous motivation are closely linked to differences in TSIR.Influence and Proximity are significant predictors of autonomous motivation (bs = .21

and .23, ps < .01 and<.001 respectively). The positive effect of Influence is equally strongfor all students. However, the effect of Proximity seems to be stronger for students in

high-ability classes, although in mixed-ability classes, its effect is notable as well. All in all

Figure 4. General trajectories of controlled and autonomous motivation during the first year of

secondary education (raw scores).

Figure 5. The developmental trajectories of controlled and autonomous motivation according to class

type based on best-fitted multilevel models.

Interpersonal relationship and academic motivation 473

Table

6.Resultsofmultilevelmodelsofthedevelopmentofautonomousmotivationovertime

Variable

Model1

a

Autonomousmotivation

Model1

b

Autonomousmotivation

withInfluence

andcontrol

variables

Model2

c

Autonomousmotivation

withProximityandcontrol

variables

Model4

d

Autonomousmotivation

withTSIRandcontrol

variables

Coefficient

SECoefficient

SECoefficient

SECoefficient

SE

Fixedeffect

Intercept

4.2558***

0.0616

4.1288***

0.0816

4.0108***

0.0809

4.0163***

0.0821

Time

�0.0264**

0.0087

�0.0189°

0.0101

�0.0246*

0.0105

�0.0248*

0.0106

Time2

Subject

taught

�0.0241

0.0602

0.0355

0.0720

0.0550

0.0702

0.0522

0.0702

Time9

subject

taught

0.0004

0.0084

�0.0076

0.0095

�0.0056

0.0100

�0.0051

0.0101

Class

type

0.1123°

0.0604

0.2673***

0.0794

0.3505***

0.0780

0.3347***

0.0801

Time9

classtype

0.0230**

0.0085

0.0031

0.0096

0.0089

0.0101

0.0078

0.0101

Studentgender

�0.0153

0.0518

�0.1193°

0.0716

�0.1142

0.0710

�0.1146

0.0711

Time9

studentgender

�0.0001

0.0074

0.0122

0.0095

0.0130

0.0094

0.0127

0.0094

Influence

0.2120**

0.0728

�0.0214

0.0845

Proximity

0.2288***

0.0366

0.2344***

0.0430

Class

type9

Influence

�0.0694

0.0890

0.0811

0.1017

Class

type9

Proximity

�0.0943*

0.0483

�0.1160*

0.0554

Random

effect

Level3

variance

(class)

Intercept

0.0040

0.0051

0.0009

0.0075

0.0001

0.0071

0.0000

0.0071

Intercept9

time

0.0009

0.0005

0.0012

0.0007

0.0012

0.0008

0.0012

0.0008

Time

0.0000

0.0001

0.0000

0.0001

0.0001

0.0001

0.0001

0.0001

Level2

variance

(student)

Intercept

0.1595

0.0219

0.2765

0.0419

0.2707

0.0413

0.2713

0.0413

Intercept9

time

�0.0090

0.0026

�0.0216

0.0051

�0.0217

0.0051

�0.0217

0.0051

Time

0.0020

0.0004

0.0032

0.0008

0.0031

0.0008

0.0031

0.0008

Level1

variance

(occasion)

Residual

0.3054

0.0116

0.2873

0.0134

0.2851

0.0133

0.2849

0.0133

Deviance

4633.81

3705.12

3666.14

3664.97

Note.

Therandomeffectsoftimeandtimequadraticareextremelysm

all(close

tozero)because

thedisplayprecisioninthemodelsisrestrictedto

fourdigitsafterthedecimalpoint.TSIR,

Teacher–studentinterpersonalrelationship.

an=2,378.bn=1,894.cn=1,894.dn=1,894.

°p<.10.*p<.05.**p

<.01.***

p<.001.

474 Ridwan Maulana et al.

therefore Proximity seems to be an important classroomenvironment characteristic for all

types of classes; it has a particularly strong effect on autonomousmotivation in high-ability

classes. Influence also plays also a role, but has a less pronounced effect. Considering that

the level of Influence remains almost the same throughout the school year (with a smalldownward trend) in high-ability classes, whereas it notably increases in mixed-ability

classes, and combining this insight with the increase in Proximity mentioned earlier, we

can conclude that the differences in the trends in Influence between the two types

constitute the ultimate cause of the fairly stable autonomous motivation of students in

mixed-ability classes, whilst leading to downward trends in autonomous motivation in

high-ability classes.

Discussion

In this study, we investigated changes in TSIR and its effects on student motivation in the

first year of secondary school in Indonesia. Contrary to the results of previous studies from

Western countries, which have either documented an overall decrease in TSIR over time

(Brekelmans, 1989; Flanders et al., 1968; Flores, 2003; Opdenakker et al., 2012; Pianta &

Stuhlman, 2004) or it remaining relatively stable (Curby et al., 2010; Jerome et al., 2008;Mainhard et al., 2011), we found that both Influence and Proximity increase in a linear

fashion over time. Existing research recognizes that higher levels of Influence and

Proximity correspond with higher levels of student motivation and better academic

achievement (den Brok et al., 2004; Maulana et al., 2011; Wubbels et al., 2006). Our

findings suggest that, in the course of the first school year, the social climate in the

Indonesian classroom becomes progressively more favourable.

Comparing our results with those found in Western countries, the more positive

perception Indonesian students have of their teachers might well stem from their culturalbeliefs, in particular the value of respecting people in positions of authority. In a

collectivist culture such as Indonesia, the power distance between student and teacher is

considerably large, reflecting the skewed power balance in society as a whole (Hofstede,

1991). Contrary to countries with smaller power differences – where authority is rather

seen as an infringement on the rights of the individual – people from cultures with a large

internal power distance tend to accept authority in everyday social interaction (Ho,

Holmes, & Cooper, 2004). In the context of the classroom, the teacher is in a position of

authority: Students are expected to acknowledge this and will see the teacher as a rolemodel. In linewith our results, another recent study on TSIR applying an SDT perspective

also showed that, in the Indonesian classroom, teacher involvement, structure, and

autonomy support increase in the course of time (Maulana & Opdenakker, 2014).

However, it should be noted that the interpretation and explanation of the findings of the

present study are tentative and limited to the context issue. Explanations of context

differences related to culture should be takenwith care. Because Indonesian andWestern

contexts differ in terms of culture as well, future research focusing on cross-cultural

studies should take a set of cross-cultural factors into account.The results of our study suggest that the level of teacher Influence remains almost

stable over time, showing just a small downward trend in high-ability classes and a slight

increase in mixed-ability classes. Levels of Proximity, on the contrary, are consistently

higher in high-ability classes. The results for Proximity (and the initial level of Influence)

concord with those of previous studies, suggesting that the learning environment is

usually more favourable in high-ability classes (Opdenakker & Van Damme, 2001;

Opdenakker, Van Damme, De Fraine, Van Landeghem, & Onghena, 2002). Perhaps

Interpersonal relationship and academic motivation 475

mixed-ability classes aremore difficult tomanage, requiring teachers to bemore dominant

and forcing them to act as leaders in order to get students to cooperate (Boufard &

Couture, 2003; Evertson, 1982). For the same reason, teachers might show more

cooperative behaviour in high-ability classes.With regard to gender differences, girls generally report higher levels of Influence than

boys. Previous studies have found that conflict, which may lead to a higher level of

controlling behaviour and strictness of teachers, is more readily accepted by boys than by

girls (Hamre & Pianta, 2001; Ladd et al., 1999; Saft & Pianta, 2001). Remarkably enough, a

recent study conducted in the Netherlands found exactly the opposite result (Opdenak-

ker et al., 2012). Perhaps cultural gender role differences can explain this discrepancy;

and teachers treat boys and girls differently in different countries (Marcus et al., 1991;

Timm, 1999). In the Netherlands, for instance, boys are often seen as troublemakers,whereas in Indonesia it is particularly important for girls to be obedient. This may cause

Dutch teachers (and probably teachers inWestern countries in general) to focus more on

boys and lead to Influence-related behaviour. Indonesian teachers, on the other hand

(and perhaps teachers in East Asian cultures in general), will behave more dominantly

towards girls.

Consistent with previous studies, students’ controlled motivation was shown to

increase in the course of the school year, with a small deceleration after 7 months

(Anderman et al., 1999; Harter, 1981; Opdenakker et al., 2012). Autonomous motiva-tion decreases linearly over time (Bouffard, Marcoux, Vezeou, & Bordeleau, 2003;

Corpus et al., 2009; Opdenakker et al., 2012; Otis et al., 2005). The steady trend of

decreasing autonomous motivation also concurs with previous research, which has

shown that students’ (intrinsic) motivation tends to wane as they grow older (Gottfried

et al., 2001; Lepper et al., 2005; Marsh, Craven, & Debus, 1998; Opdenakker et al.,

2012; Van der Werf et al., 2008). As this finding suggests, motivational problems during

the phase of the transition between schools are a common phenomenon, irrespective of

cultural setting. It has to be added, however, that when it comes to levels of studentmotivation (average mean scores) cultural differences can be said to exist. When

comparing our results with those found in a study conducted in the Netherlands

(Opdenakker et al., 2012), we can see that, in the Indonesian sample, the levels of

controlled and autonomous motivation are slightly higher. This seems to imply that the

importance of school differs between the two cultural settings: In Indonesia (and, by

extension, East Asian countries), school is considered more important than it is in

Western countries (Leung, 2002).

Consistent with the research conducted by Opdenakker et al. (2012), we also foundthat controlled motivation was slightly higher in mixed-ability classes than in high-ability

classes. Likewise, students in high-ability classes show a more rapid decline in

autonomous motivation as compared to students in mixed-ability classes. Taking into

account the differences in TSIR (changes) and the differential sensitiveness to TSIR

between the two class types, the implication appears to be that in high-ability classes –which often have a more competitive classroom climate – academic motivation develops

in a less favourable manner. Further research might shed some light on how and why the

effect of TSIR varies between the two class types; insight should be gained into howhigh-ability students can be prevented from suffering a too sharp decline in academic

motivation.

Finally,we found that (changes in) TSIR are related to both controlled and autonomous

motivation. As expected, the effects of TSIR on more autonomous forms of motivation

seem to be stronger, a finding which is consistent with earlier research (Maulana et al.,

476 Ridwan Maulana et al.

2011). In SDT, autonomous motivation occupies a higher place on the motivational

continuum, and itmightwell be that a progressive positive development of TSIR facilitates

the process of students’ internalization of learning, which will subsequently promote

their self-determined motivation for studying (Ryan & Deci, 2000). Influence andProximity are indeed important predictors of motivation, which can protect students

from becoming discouraged during the period of schooling.

Implications and future research

As our study has shown, TSIR is not stable over time. This instability not only has

consequences for students’ academic motivation, it also has far-reaching implications for

the question of how to arrive at an adequate measurement of the social climate in theclassroom (Curby et al., 2010). A limited number of measuring moments will only render

snapshots of a (highly) dynamic process, offering but little understanding of how things

develop in actual practice (Meyer, Linn, & Hastings, 1991). The finding that TSIR changes

over the course of the school year, and is longitudinally linked to changes in academic

motivation, is consistent with the findings of studies in Western countries and seems to

support that TSIR is universally important, irrespective of countries.

We derived our conceptualization of TSIR in terms of the dimensions Influence and

Proximity from theMITB framework (Wubbels et al., 2006), which overlaps theoreticallywith the operationalization in SDT of TSIR by means of the variables of Involvement,

Structure, and Autonomy support (Ryan & Deci, 2000). Having said this, the two

perspectives mentioned should still be conceived of as different theoretical view points.

Future research would benefit from identifying the extent to which the two perspectives

overlap, so as to refine the theory on TSIR.

Moreover, more attention will have to be paid to the pivotal influence of diverse

personal and contextual characteristics, in particular class type and gender. As our results

indicate, the quality of TSIR and level of student motivation, and the effects of TSIR onmotivation from a dynamic point of view, are greatly influenced by these variables. To

illustrate, our results show that, although teacher Proximity is generally higher in

high-ability classes, the increase of Influence over time is more pronounced in

mixed-ability classes. Because high levels of interpersonal Influence and Proximity are

desirable to support students’ academic outcomes (Wubbels & Levy, 1993), and research

demonstrates that the combination of interpersonal Influence and Proximity together is

important to increase academic motivation, not only one of them (Opdenakker et al.,

2012; Wubbels et al., 2006), the practical implication could be that teachers ofhigh-ability classes who wish to improve their level of Influence should take an example

after their colleagues in mixed-ability classes. Similarly, teachers of mixed-ability classes

could learn from colleagues in high-ability classes when it comes to the matter of

maintaining (and increasing) their level of Proximity. It would be potentially worthwhile

to prepare future teachers for such differences during their training and make them

develop sufficient skills to deal with and adapt to different types of classes, in order at a

higher-quality TSIR (Maulana, Helms-Lorenz, et al., 2013). As a last point to mention,

although our study comprises a relatively large number of students, this study is subject tosome limitations. First, the number of classes and teachers in our sample is fairly limited.

Second, this study is limited to the Indonesian context only. Third, the longitudinal

measurement is limited to one school year only. Fourth, this study is based heavily on

student perceptions. We therefore caution against making broad generalizations,

suggesting instead that our research should be replicated with a larger sample and in

Interpersonal relationship and academic motivation 477

different cultural contexts, with more measurement years included, and with observa-

tional report added.

Acknowledgements

We would like to specially thank the editor for his invaluable comments during the review

process and the anonymous reviewers for their constructive and useful feedback. We are

indebted to Prof. Perry den Brok for his special assistance in examining the quality of the main

measure used in this study and to Prof. Maarten Vansteenkiste for his valuable insights during

our discussion about the motivational framework and measure. We are grateful to all teachers

and students who participated in this longitudinal study and to our research assistant for her

excellent assistance during the data collection of the study. This research was conducted as

part of PhD study of the first author, while the second author received a grant from Rosalind

Franklin Fellowships (University of Groningen).

References

Anderman, E. M., Maehr, M. L., & Midgley, C. (1999). Declining motivation after the transition to

middle school: School can make a difference. Journal of Research and Development in

Education, 32, 131–147.Anderman, L. H., Patrick, H., & Ryan, A. M. (2004). Promoting adaptive motivational beliefs in

middle school classrooms. Middle School Journal, 35, 33–39.Baker, D., & Leary, R. (1995). Letting girls speak out about science. Journal of Research in Science

Teaching, 32, 3–27.Boufard, T., & Couture, N. (2003). Motivational profiles and academic achievement among students

enrolled in different schooling tracks. Educational Studies, 29, 19–38.Bouffard, T., Marcoux, M.-F., Vezeou, C., & Bordeleau, L. (2003). Changes in self-perceptions of

competence and intrinsic motivation among elementary school children. British Journal of

Educational Psychology, 73, 171–186.Brekelmans, M. (1989). Interpersonal teacher behaviour in the classroom [In Dutch:

Interpersoonlijk gedrag van docenten in de klas]. Utrecht, The Netherlands: W. C. C.

Corpus, J. H., McClintic-Gilbert, M., & Hayenga, A. (2009). Within-year changes in children’s

intrinsic and extrinsic motivational orientations: Contextual predictors and academic

outcomes. Contemporary Educational Psychology, 34, 154–166.Curby, T. W., Grimm, K. J., & Pianta, R. C. (2010). Stability and change in early childhood

classroom interactions during the first two hours of a day. Early Childhood ResearchQuarterly,

25, 373–384.Davis, H. A. (2003). Conceptualizing the role and influence of teacher-student relationships are

effective: A meta-analysis. Review of Educational Research, 77, 113–143.den Brok, P., Brekelmans, M., & Wubbels, T. (2004). Interpersonal teacher behaviour and student

outcomes. School Effectiveness and School Improvement, 15, 407–442.den Brok, P., Brekelmans, M., & Wubbels, T. (2006). Multilevel issues in studies using students’

perceptions of learning environments: The case of the questionnaire on teacher interaction.

Learning Environments Research, 9, 199–213.den Brok, P., Van Tartwijk, J., Wubbels, T., & Veldman, I. (2010). The differential effect of the

teacher-student interpersonal relationship on student outcomes for students with different

ethnic backgrounds. British Journal of Educational Psychology, 80, 199–221.Evertson, C. M. (1982). Differences in instructional activities in higher- and lower-achieving junior

high English and math classes. The Elementary School Journal, 82, 329–350.Evertson, C. M., & Veldman, D. J. (1981). Changes over time in process measures of classroom

behaviour. Journal of Educational Psychology, 73, 156–163.

478 Ridwan Maulana et al.

Feldlaufer, H., Midgley, C., & Eccles, J. S. (1988). Student, teacher, and observer perceptions of

the classroom environment before and after the transition to junior high school. Journal of

Early Adolescence, 8, 133–156.Flanders, N. A., Morrison, B. M., & Brode, E. L. (1968). Changes in pupil attitudes during the school

year. Journal of Educational Psychology, 59, 334–338.Flores, M. A. (2003).Mapping teacher change: A two-year empirical study. Paper presented at the

annual meeting of the American educational Research Association, Chicago, IL.

Fraser, B. J. (2007). Classroom learning environments. In S. K. Abell & N. G. Lederman (Eds.),

Handbook of research on science education (pp. 103–124). Mahwah, NJ: Lawrence Erlbaum.

Freeman, T. M., Anderman, L. H., & Jensen, J. M. (2007). Sense of belonging in college freshmen at

the classroom and campus levels. Journal of Experimental Education, 75, 203–220.Gottfried, A. E., Fleming, J. S., &Gottfried, A.W. (2001). Continuity of academic intrinsicmotivation

from childhood through late adolescence: A longitudinal study. Journal of Educational

Psychology, 93, 3–13.Hambleton, R. K. (1994). Guidelines for adapting educational and psychological tests: A progress

report. European Journal of Psychological Assessment, 10, 229–244.Hamre, B.K., &Pianta, R. C. (2001). Early teacher-child relationships and the trajectory of children’s’

school outcomes through eighth grade. Child Development, 72, 625–638.Harter, S. (1981). A model of mastery motivation in children: Individual differences and

developmental change. In W. A. Collins (Ed.), The Minnesota symposia on child psychology:

Aspects of the development of competence (Vol. 14, pp. 215–255). Hilsdale, NJ: Lawrence

Erlbaum.

Henderson, D., Fisher, D. L., & Fraser, B. J. (2000). Interpersonal behaviour, laboratory learning

environments, and student outcomes in senior biology classes. Journal of Research in Science

Teaching, 37, 26–43.Ho, E., Holmes, P., & Cooper, J. (2004). Review and evaluation of international literature on

managing cultural diversity in the classroom. Hamilton, New Zealand: University of Waikato.

Hofstede, G. (1991).Cultures and organizations: Software of themind: Intercultural cooperation

and its important for survival. New York, NY: McGraw-Hill.

Huang, S.-Y. L., & Waxman, H. C. (1992). Stability of teachers’ classroom instruction across

classes and time of observation. Paper presented at the Annual Meeting of the American

Educational Research Association, San Francisco, CA.

Ireson, J., & Hallam, S. (2009). Academic self-concepts in adolescence: Relations with achievement

and ability grouping in schools. Learning and Instruction, 19, 201–213. doi:10.1016/j.

learninstruc.2008.04.001

Jerome, E. M., Hamre, B. K., & Pianta, R. C. (2008). Teacher-child relationships from kindergarten to

sixth grade: Early childhood predictors of teacher-perceived conflict and closeness. Social

Development, 18, 915–945.Kahle, J. B., Parker, L. H., Rennie, L. J., & Riley, D. (1993). Gender differences in science education:

Building a model. Educational Psychologist, 28, 379–404. doi:10.1207/s15326985ep2804_6Ladd, G. W., Birch, S. H., & Buhs, E. S. (1999). Children’s social and scholastic lives in kindergarten:

Related spheres of influence? Child Development, 70, 1373–1400.Lapointe, J. M., Legault, F., & Batiste, S. J. (2005). Teacher interpersonal behaviour and adolescents’

motivation in mathematics: A comparison of learning disabled, average, and talented students.

International Journal of Educational Research, 43, 39–54.Leary, T. (1957).An interpersonal diagnosis of personality. NewYork,NY: Ronald Press Company.

Lepper, M. R., Corpus, J. H., & Iyengar, S. S. (2005). Intrinsic and extrinsic motivational orientations

in the classroom: Age differences and academic correlates. Journal of Educational Psychology,

97, 184–196.Leung, F. K. S. (2002). In search for East Asian identity in mathematics education. Educational

Studies in Mathematics, 47, 35–51.Levy, J., den Brok, P., Wubbels, T., & Brekelmans, M. (2003). Students’ perceptions of interpersonal

aspects of the learning environment. Learning Environments Research, 6, 5–36.

Interpersonal relationship and academic motivation 479

Liem, G. A. D., Martin, A. J., Nair, E., Bernardo, A. B. I., & Prasetya, P. H. (2009). Cultural factors

relevant to secondary school students in Australia, Singapore, Philippines and Indonesia:

Relative differences and congruencies. Australian Journal of Guidance and Counselling, 19

(2), 161–178.Mainhard, M. T., Brekelmans, M., den Brok, P., & Wubbels, T. (2011). The development of the

classroom social climate during the first months of the school year. Contemporary Educational

Psychology, 36, 190–200.Marcus, G., Gross, S., & Seefeldt, C. (1991). Black and White students’ perceptions of teacher

treatment. The Journal of Educational Research, 84, 363–367.Marsh, H. W., Craven, R., & Debus, R. (1998). Structure, stability, and development of young

children’s self-concept: A multi-cohort multi-occasion study. Child Development, 69, 1030–1053.

Maulana, R., Helms-Lorenz, M., & Van de Grift, W. (2013). Development and evaluation of a

questionnaire measuring pre-service teachers’ teaching behaviour: A Rasch modelling

approach. School Effectiveness and School Improvement. Advance online publication.

Maulana, R., & Opdenakker, M.-C. (2014). Do teacher–student relationships deteriorate over time?

An investigation of within-year changes and links with autonomous motivation in Indonesia. In

T. Wubbels, P. den Brok, J. van Tartwijk & J. Levy (Eds.), Advances in Learning Environments

Research (pp. 133–158). Rotterdam, the Netherlands: Sense Publishers.

Maulana, R., Opdenakker, M.-C., den Brok, P., & Bosker, R. (2011). Teacher-student

interpersonal relationships in Indonesia: Profiles and importance to student motivation. Asia

Pacific Journal of Education, 31, 33–49.Maulana, R., Opdenakker, M.-C., den Brok, P., & Bosker, R. (2012a). Teacher-student interpersonal

relationships during the first year of secondary education: A multilevel growth curve analysis.

In T. Wubbels, P. den Brok, J. van Tartwijk & J. Levy (Eds.), Interpersonal relationships in

education: An overview of contemporary research (pp. 207–224). Rotterdam, The

Netherlands: Sense Publisher.

Maulana, R., Opdenakker, M.-C., den Brok, P., & Bosker, R. (2012b). Teacher-student interpersonal

relationships in Indonesian lower secondary education: Teacher and student perceptions.

Learning Environments Research, 15 (2), 251–271.Maulana, R., Opdenakker, M.-C., Stroet, K., & Bosker, R. (2013). Changes in teachers’ involvement

versus rejection and linkswith academicmotivation during the first year of secondary education:

A multilevel growth curve analysis. Journal of Youth and Adolescence. Advance online

publication. doi:10.1007/s10964-013-9921-9

Meece, J. L., & Miller, S. D. (2001). A longitudinal analysis of elementary school students’

achievement goals in literacy activities. Contemporary Educational Psychology, 26, 454–480.

Meyer, L. A., Linn, R. L., & Hastings, C. N. (1991). Teacher stability from morning to afternoon and

from year to year. American Educational Research Journal, 28, 825–847.Midgley, C., Kaplan, A., & Middleton, M. J. (2001). Performance–approach goals: Good for what,

for whom, under what circumstances, and at what cost? Journal of Educational Psychology,

93, 77–86. doi:10.1037/0022-0663.93.1.77Mills, R. (1997). Grouping students for instruction: Issues of equity and effectiveness. In J. L. Irvin

(Ed.),What current research says to themiddle level practitioner (pp. 87–94). Columbus, OH:

National Middle School Association.

Ministry of National Education (2007). Guidelines for establishing internationally standard

pioneering primary and secondary education. Jakarta, Indonesia: Ministry of National

Education (In Indonesian).

Mullis, I. V. S., Martin, M.O., Olson, J. F., Berger, D. R., Milne, D., & Stanco, G.M. (2008). TIMSS 2007

encyclopaedia: A guide tomathematics and science education around theworld. Boston, MA:

TIMSS & PIRLS International Study Centre.

Murphy, P. K., & Alexaneder, P. A. (2000). A motivated exploration of motivation technology.

Contemporary Educational Psychology, 25, 3–53.

480 Ridwan Maulana et al.

Muthen, L. K., & Muthen, B. O. (1999). Mplus users’ guide: The comprehensive modelling

program for applied researchers. Los Angeles, CA: Muthen and Muthen.

Newberry, M., & Davis, H. A. (2008). The role of elementary teachers’ conceptions of closeness to

students on their differential behaviour in the classroom. Teaching and Teacher Education, 24,

1965–1985.Opdenakker, M.-C., & Maulana, R. (2010). Teacher-student relationships and academic

engagement: How do they develop and link? Paper presented at the International

Conference on Interpersonal Relationships in Education, Boulder, CO.

Opdenakker, M.-C., Maulana, R., & den Brok, P. (2012). Teacher-student interpersonal relationships

and academic motivation within one school year: Developmental changes and linkage.

School Effectiveness and School Improvement, 23, 95–119.Opdenakker, M.-C., & Van Damme, J. (2001). Relationship between school composition and

characteristics of school process and their effect on mathematics achievement. British

Educational Research Journal, 27, 407–432.Opdenakker, M.-C., Van Damme, J., De Fraine, B., Van Landeghem, G., & Onghena, P. (2002). The

effect of schools and classes on mathematics achievement. School Effectiveness and School

Improvement, 13, 399–427.Otis, N., Grouzet, F. M. E., & Pelletier, L. G. (2005). Latent motivational change in an academic

setting: A 3-year longitudinal study. Journal of Educational Psychology, 97, 170–183.Patrick, H., Turner, J. C., Meyer, D. K., & Midgley, C. (2003). How teachers establish psychosocial

environments during the first days of school: Associations with avoidance in mathematics.

Teachers College Record, 105 (8), 1521–1558.Pianta, R. C., & Stuhlman, M. W. (2004). Teacher–child relationships and children’s success in the

first years of school. School Psychology Review, 33, 444–458.Rasbash, J., Steele, F., Browne, W. J., & Goldstein, H. (2009). A user’s guide to MLwiN. Bristol, UK:

Centre for Multilevel Modelling, University of Bristol.

Ratelle, C. F., Guay, F., Larose, S., & Senecal, C. (2004). Family correlates of trajectories of academic

motivation during a school transition: A semi parametric group-based approach. Journal of

Educational Psychology, 96, 743–754.Ryan, R. M., & Connell, J. P. (1989). Perceived locus of causality and internalization: Examining

reasons for acting in two domains. Journal of Personality and Social Psychology, 57, 749–761.Ryan, R. M., & Deci, E. L. (2000). Self-determination theory and the facilitation of intrinsic

motivation, social development, and well-being. The American Psychologist, 55, 68–78.Ryan, A. M., & Patrick, H. (2001). The classroom social environment and changes in adolescents’

motivation and engagement during middle school. American Educational Research Journal,

38, 437–460.Saft, E. W., & Pianta, R. C. (2001). Teacher’s perceptions of their relationships with students:

Relations with child and teacher characteristics. School Psychology Quarterly, 16, 125–141.Skinner, E. A., & Belmont, M. J. (1993). Motivation in the classroom: Reciprocal effects of teacher

behaviour and student engagement across the school year. Journal of Educational Psychology,

85, 571–581.Snijders, T. A. B., & Bosker, R. J. (1999).Multilevel analysis: An introduction to basic and advance

multilevel modelling. London, UK: Sage.

Spinath, B., & Spinath, F. M. (2005). Longitudinal analysis of the link between learning motivation

and competence beliefs among elementary school children. Learning and Instruction, 15, 87–102.

Timm, J. T. (1999). The relationship between culture and cognitive style: A review of the evidence

and some reflections for the classroom. Educational Researcher, 28, 36–44.Triandis, H. C. (1993). Culture and social behaviour. London, UK: McGraw-Hill.

Turner, J. C., & Patrick, H. (2004). Motivational influences on student participation in classroom

learning activities. Teachers College Record, 106, 1759–1785.Usher, E. L., & Pajares, F. (2009). Sources of self-efficacy in mathematics: A validation study.

Contemporary Educational Psychology, 34, 89–101. doi:10.1016/j.cedpsych.2008.09.002.

Interpersonal relationship and academic motivation 481

Van de Gaer, E., De Fraine, B., Pustjens, H., Van Damme, J., De Munter, A., & Onghena, P. (2007).

School effects on the development of motivation toward learning tasks and the development

of academic self-concept in secondary education: A multivariate latent growth curve

approach. Paper presented at the Annual Meeting of the American Educational Research

Association, Chicago, IL.

Van derWerf, G., Opdenakker, M.-C., & Kuyper, H. (2008). Testing a dynamic model of student and

school effectiveness with a multivariate multilevel latent growth curve approach. School

Effectiveness and School Improvement, 19, 447–462.Vansteenkiste, M., Simons, J., Lens, W., Sheldon, K. M., & Deci, E. L. (2004). Motivating learning,

performance, and persistence: The synergistic role of intrinsic goals and autonomy-supportive

contexts. Journal of Personality and Social Psychology, 87, 246–260.Vansteenkiste, M., Zhou, M., Lens, W., & Soenens, B. (2005). Experiences of autonomy and control

among Chinese learners: Vitalizing or immobilizing? Journal of Educational Psychology, 97,

468–483.Watt, H. M. G. (2008). A latent growth curve modeling approach using an accelerated

longitudinal design: The ontogeny of boys’ and girls’ talent perceptions and intrinsic values

through adolescence. Educational Research and Evaluation, 14, 287–304. doi:10.1080/

13803610802249316

Wentzel, K. R. (2002). Are effective teachers like good parents? Teaching styles and student

adjustment in early adolescence. Child Development, 73, 287–301.Wentzel, K. (2010). Teacher-student relationships and classroom supports: Understanding

motivation and engagement in young adolescents. Paper presented at the International

Conference on Interpersonal Relationships in Education, Boulder, CO.

Wubbels, T., & Brekelmans, M. (2005). Two decades of research on teacher-student relationships in

class. International Journal of Educational Research, 43, 6–24.Wubbels, T., Brekelmans, M., den Brok, P., & van Tartwijk, J. (2006). An interpersonal perspective

on classroom management in secondary classrooms in the Netherlands. In C. Evertson &

C. Weinstein (Eds.), Handbook of classroom management: Research, practice, and

contemporary issues (pp. 1161–1191). New York, NY: Lawrence Erlbaum Associates.

Wubbels, T., & Levy, J. (1993). Do you know what you look like?: Interpersonal relationships

in education. London, UK: Falmer Press.

Yeung, A. S., Lau, S., & Nie, Y. (2011). Primary and secondary students’ motivation in learning

English: Grade and gender differences. Contemporary Educational Psychology, 36, 246–256.

Received 31 July 2012; revised version received 10 August 2013

482 Ridwan Maulana et al.