Research in Science Education

, Volume 45, Issue 4, pp 509–526 | Cite as

An Expectancy-Value Model for Sustained Enrolment Intentions of Senior Secondary Physics Students



This study investigates the predictive influences of achievement motivational variables that may sustain students’ engagement in physics and influence their future enrolment plans in the subject. Unlike most studies attempting to address the decline of physics enrolments through capturing students’ intention to enrol in physics before ever studying the subject, this study is novel because it captures the perceptions of students currently enrolled in senior secondary physics and their subsequent enrolment intentions after completing modules from the physics curriculum. Participants comprised of senior secondary students in year 11 completing their first year of physics in Australia across nine high schools in New South Wales. The Sustained Enrolment Models for Physics (SEMP), which drew upon the Expectancy-Value (EV) theoretical foundation, proposed predictive relations among students’ achievement motivation, sustained engagement, and enrolment intentions in relation to physics. The data showed a good fit to the theoretically developed model for all four physics topics from the year 11 curriculum. The path coefficients of the models demonstrated the strength of relationships among the variables for each of the topics. The topic specificity of SEMPs allowed the mapping of students’ motivational patterns at a more sensitive level than the domain-specific level and suggested that the relative influence of motivational precursors can vary by topic. This study advanced the EV research knowledge that, while values may be significant, it is the expectancies that largely predict students’ sustained choice intentions in relation to physics. Implications for these findings are discussed.


Physics education Motivation and engagement with physics Sustained enrolment plans Expectancy-Value theory Senior secondary physics students 


Across decades, there has been increasing international research interest in the declining physics enrolment trend in schools and higher education institutions. This trend is disconcerting since it is a priority to raise the level of scientific literacy of our society (Osborne et al. 2003), and with the projected increase in jobs over the next decade within this field (e.g., STEM; Education for Global Leadership n. d.) particularly in industrialized countries including Australia, it is critical we prepare a workforce to fill these positions. Studies investigating the decline in enrolments have been conducted in Australia (e.g., Ainley et al. 2008; Barnes 1999; Goodrum et al. 2001; Lyons and Quinn 2010), Europe, and the USA (e.g., Bennett et al. 2013; Gill and Bell 2013; Krogh and Thomsen 2005). Similar concerns have been reported from several other countries such as Malaysia (Khalijah 2004), Japan (Goto 2001), and India (Garg and Gupta 2003). Prompted by this alarming trend researchers, including those referred to above, have examined the motivational and social factors that are predicted to influence students’ choices to either enrol or not in physics.

Physics is characterised as a subject that has a significant strategic value for entry into prestigious courses that can lead to high status career paths (Fullarton and Ainley 2000; Lyons 2006). Across the international literature, this instrumental value of physics is the most influential factor which predicts students’ physics enrolment plans (e.g., Barnes 1999; Eccles et al. 1998; Mujtaba and Reiss 2013; Stokking 2000). Specifically in Australia, perceived instrumental value of physics has been suggested as the largest predictor of physics enrolment in senior secondary schools (e.g., Barnes 1999; Woods 2008), and any decline in this value is found to negatively influence the enrolment pattern (Lyons and Quinn 2010). Likewise, the perceived difficulty level of the subject is another major negative influence on student enrolment plans (Spall et al. 2003). For females, another competing factor influencing their decision to study physics is the gender-biased stereotypes affiliated with the subject such that it is perceived to be a course and career path that is not as suitable for females as it is for males (Bultitude et al. 2010; Mujtaba and Reiss 2013). Gill and Bell (2013) reported that even among high achieving students, there exists the widely held belief that physics is a subject for boys, and this prevailing attitude contributes to lower participation of females in this subject. Furthermore, research has shown a lack of gender inclusivity of the curriculum which reinforces the gender-biased affiliations with the subject (Goodrum et al. 2001; Rennie and Parker 1996) and hence negatively influences female students’ decision to enrol in physics. Likewise, the learning environment in a physics classroom (Krogh and Thomsen 2005) and students’ achievement in related domains such as mathematics (Ainley et al. 2008) have been other important factors influencing the choice to study physics at the senior secondary level.

While there is an increasing body of research on physics enrolment, it tends to focus on the variables that act as predictors of initial intentions to enrol in physics in the future. The participants in these studies were those students who had just completed a preparatory course in general sciences prior to enrolling in an elective science at the senior secondary level. This implies that while the existing research has supplied a rich interpretation of the interactions among various motivational constructs affecting students’ initial enrolment intentions in physics, it remains unclear what factors sustain their intention of further physics enrolment after actually encountering the enacted physics curriculum at the senior secondary level. Instead of focusing exclusively on future enrolment intentions, perhaps we could learn new insights into studying students’ motivation while actually undertaking senior secondary science. Through understanding the key variables that sustain students’ motivation in science, we may maintain student engagement across time leading to increases in students choosing the subject at school but also considering it as a career pathway into the future.

In Australia, we have a unique exit point from studying physics which presents the opportunity to examine sustained students’ enrolment intentions. Specifically, Australia has a 2-year senior high school structure, where students can decide at the end of first year to discontinue or continue physics (provided they satisfy a total minimum of 10 units of study) to the final year. That is, students who have shown an initial motivation to study physics can commence their study in year 11, and after an extended period of time (end of Year 11), they are given the opportunity to sustain or abandon their further enrolment plans going into their final year of school (year 12). This structure allows for the opportunity to examine whether their initial motivation is sustained or decreased during the academic year after encountering the curriculum. Consequently, the relative strengths of motivational variables that contribute to physics engagement can be identified.

Aim of the Study

The purpose of this study is to examine students’ sustained enrolment intentions in physics based on Expectancy-Value (EV) theory of achievement motivation. To do this, we tested a model that was theoretically informed by the General Model of Academic Choice (GMAC) (Eccles et al. 1983) and adapted for a topic-specific level with a unique focus on examining sustained motivation for senior secondary students. This model is referred to hereon in as the Sustained Enrolment Model for Physics (SEMP).

Previous studies have tended to measure motivational patterns in physics at the subject area or domain level (e.g., Barnes 1999; Woods 2008). This domain-specific approach treats physics as a unit of study whereby motivational patterns are considered to be stable across the different topics within the domain. This measurement approach however overlooks what is actually occurring within the physics classroom. Importantly, the physics curricula at higher education levels, including the senior secondary years, consists of different topics with varying characteristics such that some topics are descriptive, while some are more problem oriented. Some are theoretical, while others have more practical utility for everyday life. Students may exhibit quite different behaviors and attitudes towards the different topics. The possibility that the students’ motivational patterns in physics may vary across topics appears to have been overlooked in physics motivation studies. The present study accounts for this potential topic-specific influence by measuring students’ motivation and enrolment intentions at the completion of each of their physics topics.

Sustained Enrolment Model for Physics

For senior secondary school students, the initial motivation to enrol in physics, a subject that they have not previously studied, does not equate to sustained engagement with the subject. Some students may actually become less engaged with a subject as they become more familiar with it. Therefore, SEMP was developed on an integrated theoretical framework where the key predictor variables from EV theory were combined with sustained engagement. Both motivation and engagement are crucial in enhancing learning and persisting in a task, and are reciprocally related (Eccles 2008). Although student motivation in learning is often inferred from engagement with the learning tasks or subjects, arguably they are distinct. Ainley (2004) distinguishes between these constructs, proposing, “motivation is about energy and direction, the reasons for behavior, why we do what we do. Engagement describes energy in action; the connection between person and activity” (p. 2). Therefore, the SEMPs examined the mediational role of sustained engagement with physics on the sustained enrolment intentions of physics students.

Predictors of Physics Enrolment

An extensive search of physics enrolment literature on EV theoretical perspectives indicates that any model attempting to explain physics enrolment motivation should focus on the (1) interest value, (2) utility value, (3) performance perceptions, (4) sex-stereotyped attitudes, and (5) engagement. Consequently, the SEMP model included these variables because they would likely influence students’ sustained engagement with physics, which in turn would predict their sustained enrolment intentions in physics.



The participants were year 11 physics students from nine high schools (Government and Catholic) in New South Wales, Australia. Data were collected using the Physics Motivation Questionnaire (PMQ) measuring the constructs of the SEMP on four occasions during the 2009 academic year (See appendix 1 for PMQ and its psychometric properties). The PMQ comprises of 22 items measuring six constructs on a six-point Likert scale (1 = strongly disagree and 6 = strongly agree). The development and testing of this questionnaire are described elsewhere (Abraham and Barker in press).

Material and Procedure

There are four modules in the NSW year 11 physics curriculum, namely: The World Communicates (commonly referred to as waves); Moving About (motion); Electrical Energy at Home (electricity); and The Cosmic Engine (cosmic engine). The four time points of data collection corresponded to completion of each of the four physics modules (topics). Items of PMQ were made module-specific at each time point. The questionnaires were administered in the classrooms at the completion of each module under the supervision of the classroom physics teacher.

The total sample size across the nine schools varied across the four time points (T1 = 270, T2 = 280, T3 = 239, and T4 = 222). This variation was mainly due to student attendance on the day the survey was administered. The sample across all four data collection points remained higher for males than females, as female participation in physics is significantly lower than that of males in NSW senior secondary physics (Fullarton et al. 2003). While the numbers of male students at the four time points were 178, 180, 147, and 140, the numbers of females were 92, 100, 92, and 82, respectively.

Interest Value

Interest value (interest) refers to the “inherent enjoyment or pleasure one gets from engaging in an activity” (Eccles et al. 2005, p. 239). In this study, the interest of a physics module was related to the goals such as having a real desire to study more about the particular physics module and looking forward to learning more about the particular module.

Utility Value

Utility value (utility) is defined as the “value a task acquires because it is instrumental in reaching a variety of long-and short-range goals” (Eccles and Wigfield 1995, p. 216). In this investigation, the utility of a particular physics module was related to students’ goals, such as securing a good job, gaining entry to a future career (e.g., Engineering), and pursuing particular study plans at university or higher education institutions.

Performance Perceptions

The task-specific beliefs identified for this investigation are labeled as students’ performance perceptions (perfperc) in that particular physics module. This construct subsumes two highly related constructs: namely, students’ self-concept of ability in the particular module and their perceptions of task difficulty of the physics module. Eccles et al. (1983) demonstrated that self-concept of ability and perceptions of task difficulty interact in predicting expectations for success in a subject (see also Eccles and Wigfield 1995). Therefore, in this investigation, both constructs together are represented by a single construct—namely, perfperc—in the physics module.

Sex-Stereotyped Attitudes

Gender role beliefs (Eccles et al. 1983) in the particular physics module, which are a measure of the extent to which an individual believes that the particular physics module is a male domain, are included in the present model. The term sex-stereotyped attitude (sexstereo) towards a particular physics module represents the construct in this investigation. This construct is included to measure whether students’ gender role beliefs (such as males find a particular module naturally easier than females, males are naturally more interested in a module, or males can perform better in a module) continue to influence sustained engagement and further enrolment intentions in senior secondary physics classes.

Sustained Engagement

One of the outcome variables of SEMP was students’ sustained engagement with the physics module (engage) that was hypothesised to be predicted by the four EV variables. Engage was chosen as an outcome variable since its measurement captures students’ energy and commitment across time for each of the physics topics. Students’ initial motivation to enrol in a subject they have not previously studied is different from their engagement levels as they continue to study the subject. For instance, some students who were initially highly motivated to choose a subject may become increasingly less engaged with the subject as they become more familiar with it. Therefore, Russell et al. (2003) argued that motivation and engagement can be distinguished in a practical sense: that is, among students who are motivated to study a subject, some may be more engaged than others. Consequently, the engage outcome measure is utilised in this study to examine the mediational role of sustained engagement with physics on sustained enrolment intentions in physics. Engage was measured as a broad construct that is subsumed by interrelated aspects of behavior, emotion, and cognition in the year 11 physics module (Willms 2003).

Sustained Enrolment Intentions in Physics

The second outcome measure of the model was the students’ sustained intention to choose further physics (choicein) at the completion of a specific physics module. This variable was hypothesised to be influenced by the four EV variables and by sustained engagement (engage) at the completion of each of the four physics modules.

The relationships among the variables are depicted pictorially as a structural equation path model. The conceptual diagram for SEMP is presented in Fig. 1. The strength of the relations among the variables was inferred from the path coefficients among the constructs of SEMP and was assessed using the techniques of structural equation modeling (SEM) (Kline 1998) with LISREL 8.72 software (Jöreskog and Sörbom 1996).
Fig. 1

Conceptual diagram of the hypothesised model for sustained enrolment intentions in physics (SEMP).

Data Screening

Data screening and Cronbach’s alpha reliability estimates were achieved using SPSS 15.0. The Expectation Maximization (EM) algorithm (Schafer and Graham 2002) was employed to deal with missing data. The SEMPs depicting structural relations were generated by SEM using LISREL 8.72.

Assessing Model fit

This study employed the following criteria for examining model fit. The χ2/df ratio less than 3, the Root Mean Square Error of Approximation (RMSEA) less than or equal to 0.05, and the comparative fit indices—namely, the Comparative Fit Index (CFI) and the Tucker-Lewis Index (TLI) values greater than 0.90 were considered as an excellent model fit to the data (Byrne 1998; Kline 1998). For RMSEA, values near 0.08 indicate a fair fit, and values above 0.10 indicate a poor fit (Byrne 1998).

The significance placed by the participants on a particular construct when making enrolment decisions was demonstrated through the path coefficient of SEMPs. The structural part of the SEMP is the mathematical representation that specifies the strength and the direction in which constructs influence each other.


The mean values of the SEMP constructs measured senior secondary students’ feelings about the four achievement motivational variables, sustained engagement, and sustained enrolment intentions in relation to each of the physics topics. Given that parallel wordings were used for indicators that measure the same constructs across the four modules, it was reasonable to assume that the scores are comparable. Participants reported high or near average values for all variables (mean of the construct = 3.5) except for sexstereo, across all physics modules (see Table 1).
Table 1

Mean values of constructs across the four physics modules

Latent variables





Cosmic engine































interest = interest value of the physics module, perfperc = performance perceptions for the module, sexstereo = sex-stereotyped attitudes to the module, utility = utility value of the module, engage = sustained engagement with the module, choicein = sustained intention to continue in physics.

Sustained Enrolment Model for the Waves Module

The data for the waves module showed fair fit indices (χ2/df = 1.90, TLI = 0.958, CFI = 0.964, and RMSEA = 0.056) to the hypothesised SEMP. However, in model building, the path coefficients (the beta value and the gamma values) are considered more crucial in determining model validity (Kline 1998). The path coefficients of the model were tested for significance (Fig. 2). The significance levels were set at 0.01 levels (i.e., T values of both beta and gamma should be greater than 2.58). Moreover, following Kline’s (1998) guidelines, deleting nonsignificant paths from structural equation models strictly on empirical criteria can cause either Type 1 or Type 2 errors, therefore a greater role was given to EV theory in the model building process. Therefore, the nonsignificant paths for interest and sexstereo were retained in the SEMP for the waves module. Despite being nonsignificant for this particular topic, they are a strong determinant of enrolment choice which is consistent with EV theory.
Fig. 2

Sustained Enrolment Model for the waves module. (** = significant at 0.01)

The model demonstrated that engage can increase by 47 standard deviations given a change in perfperc of plus one standard deviation when the other EV variables are controlled for. At the same time, engage is expected to increase by 29 standard deviations given a change in utility of plus one standard deviation, when other EV variables are held constant. Engage has a significant influence on choicein, increasing it by 79 standard deviations given a change of plus one standard deviation on itself. The remaining variables were nonsignificant in their influences on other variables. The negative relation between sexstereo and engage was proposed in the hypothesised model.

This model explains the variance of the endogenous variables to a “large” extent (Hills 2008). The four EV variables together explain 32 % of the variance of engage and the model explains 64 % of the variance of choicein. Among the EV predictors, perfperc in the module was found to be the dominant predictor (explaining 14 % of the variance) of engage. This was followed by the perceived utility for the topic, which explained 3 % of the variance of engage. Both interest of the topic (the third strongest influence, explaining 1 % of the variance of engage) and sexstereo (explaining 0.4 % of the variance) were found to be nonsignificant influences on engage for the waves module. Engage explained the variance of choicein to a significant extent (64 %). Perfperc and the utility of the module were found to be significant indirect influences mediated by engage on the choicein for the waves module.

Sustained Enrolment Model for the Electricity Module

The fit indices of the SEMP for electricity indicated a fair fit (χ2/df = 2.44, TLI = 0.943, CFI = 0.951, and RMSEA = 0.072) to the data (Fig. 3). The model also explained the variance of the endogenous variables effectively (33 % of the variance of engage and 65 % of the variance of choicein were explained).
Fig. 3

Sustained Enrolment Model for the Electricity Module. (** = significant at 0.01)

Significant paths were found between engage and choicein (0.81), perfperc and engage (0.49), and between sexstereo and engage (−0.31). These were found to be significant at the 0.01 level (T values were 5.401, 4.623, and −4.064, respectively). The negative value of the sexstereo path demonstrates that engage is predicted to decline by 0.31 standard deviations given a change in sexstereo and no change in any of the other EV variables. Engage can increase by 47 standard deviations, given a change in perfperc of plus one standard deviation when the other EV variables are controlled for. At the same time, engage is expected to decrease by 31 standard deviations given a change in sexstereo of plus one standard deviation, when other EV variables are held constant. Engage has a significant influence on choicein, increasing it by 81 standard deviations given a change of plus one standard deviation. The two paths from interest and utility to engage were found to be nonsignificant at 0.01 level (T values were 0.273 and 1.034, respectively) on engage with this module.

This model explained 65 % of the variance of choicein in physics. Similarly to the previous module, perfperc was found to be the strongest predictor (explaining 8 % of variance) of engage. This was followed by sexstereo (explaining 1.4 % of the variance). Similarly to what was observed earlier for the waves module, interest was found to be nonsignificant (explaining only 0.1 % of the variance). Likewise, utility explained only 0.5 % of the variance and was identified as a nonsignificant influence on engage. Perfperc and sexstereo of the module were found to be significant indirect influences on choicein for this module.

Sustained Enrolment Model for the Motion Module

The SEMP for motion displayed fair fit indices (χ2/df = 2.27, TLI = 0.948, CFI = 0.955, and RMSEA = 0.074) to the data (Fig. 4). The model explained the variances of the endogenous variables to a large extent (explaining 42 % of the variance of engage and 69 % of the variance of choicein).
Fig. 4

Sustained enrolment model for the motion module. (** = significant at 0.01)

Significance testing of the regression paths displayed that the paths between engage and choicein (0.83), between perfperc and engage (0.44), and between sexstereo and engage (−0.32) were significant at the 0.01 level (T values being 7.314, 4.095, and −4.602, respectively). The path sizes of utility (T value = 1.777) and that of interest to engage were nonsignificant at this level (T value = 0.439). The model predicted that engage would increase by 44 standard deviations given a change in perfperc of plus one standard deviation, when the other EV variables are controlled for. Engage was expected to decrease by 32 standard deviations given a change in sexstereo of plus one standard deviation, when other EV variables are held constant. Engage has a significant influence on choicein, increasing it by 83 standard deviations given a change of plus one standard deviation.

As in the case of the previous modules, perfperc was found to have the strongest influence on engage for the module motion (explaining 12 % of the variance of engage). This was followed by sexstereo (explaining 3 % of the variance). It was also observed that, just as for the previous modules, the influence of interest was nonsignificant (explaining only 2 % of the variance) on engage. Likewise, utility explained only 1 % of the variance of engage. Perfperc and sexstereo for the module were found to have significant indirect influences on choicein in physics.

Sustained Enrolment Model for the Cosmic Engine Module

The data for Module 4 displayed a fair fit (χ2/df = 2.90, TLI = 0.929, CFI = 0.939, and RMSEA = 0.093) to the hypothesised model (see Fig. 5) and explained the variances of the endogenous variables efficiently. Thirty Six and 59 % of the variances of the two endogenous variables (engage and choicein, respectively) were explained by the model.
Fig. 5

Sustained enrolment model for the cosmic engine module. (** = significant at 0.01)

The paths between engage and choicein (0.74), between perfperc and engage (.44), and between sexstereo and engage (−.42) were all significant at the 0.01 level. The model predicted an increase by 44 standard deviations for engage given a change in perfperc of plus one standard deviation, and a decrease by 42 standard deviations, given a change in sexstereo of plus one standard deviation, when other EV variables are held constant. Engage had a significant influence on choicein, increasing it by 74 standard deviations, given a change of plus one standard deviation. The paths between interest and engage and between utility and engage were nonsignificant at this level (T values were −0.934 and 1.707, respectively).

It was found that perfperc and sexstereo significantly influenced (explaining 7 and 6 % of variances, respectively) engage for the cosmic engine. Interest was not significant (explaining 1 % of the variance of engage). Similarly, utility explained only 1 % of the variance of engage. Significant positive associations were found between the expectancy and value constructs for this module also. Perfperc and sexstereo for the module were found to have significant indirect influences on choicein.


The results established empirical support for the theoretically developed SEMP for all four physics modules. The path coefficients provided an in-depth understanding of the significance of certain factors that sustain students’ engagement (engage) and influence their decision to remain in physics (choicein). The module specificity of the models provided more sensitive measurement to examine senior secondary students’ motivation and engagement across four topics and whether these topics influenced their future enrolment plans in physics.

The uniqueness of this study lies in its endeavor to illuminate which EV theory constructs are most instrumental to senior secondary students’ physics sustained engagement and enrolment intentions. Understanding the salience of these influential motivational constructs at a topic-specific level can inform educators to ensure best practice occurs in physics classrooms and lead to increases in students’ engagement and choice to continue studying physics. Notably, the module-specific SEMPs in this study indicate that performance perceptions was the strongest determinant influencing students’ sustained engagement and this was instrumental to their choice to continue studying physics. The next strongest determinants were sustained engagement and sex stereotyped attitudes for three modules and utility value for one module.

Significant positive correlations were found between interest value and performance perceptions and between performance perceptions and utility value. This supports the positive association between expectancy and value constructs proposed by modern EV theorists (e.g., Eccles 1989; Eccles et al. 1983; Feather 1988; Meece et al. 1990; Wigfield and Eccles 1989, 1992, 2000). Across all models, it was found that sustained engagement is a strong predictor of sustained choice intentions. The majority of the variance of sustained engagement for the modules remained unexplained by the four EV motivational variables together. This is reasonable, considering the contextual nature of the engagement construct. Research suggests that student engagement with learning is highly influenced by various other factors, such as school, teacher, family, and classroom (Russell et al. n.d).

The results found in this study were different from those found by earlier studies on the EV theoretical framework, which suggest that value constructs are stronger predictors of choice than expectancy constructs. For example, in Barnes’ physics enrolment model (1999), the utility of physics, a value variable, was the largest predictor of future enrolment plans in relation to physics. Likewise, in Eccles’ models, task values predicted choices in relation to studying mathematics and physics (Eccles 1987; Eccles et al. 1983; Eccles 1984; Meece et al. 1990). However, all of these studies focused on initial enrolment intentions of students who have never done physics before and are making enrolment plans in relation to a future study of physics. In contrast, the present study focused on the factors that influence students’ intention to continue their enrolment in an enacted physics curriculum; where such intentions are shaped by their actual experience of the enacted physics curriculum. Furthermore, this study examines intentions across four distinct physics modules, while the previous studies examined enrolment motivation at a domain general level. As the current investigation is among the first to use an EV framework to explore intentions to continue on with a subject, there are no similar studies on physics enrolment available for a strict comparison of the results. A related study would be that of Hipkins et al. (2006), where New Zealand senior high school students cited a complex interplay of various factors such as career considerations, personal interest, and other strategic reasons as influential on their decision for choosing sciences including physics in year 13 (final year of senior secondary school), while their intentions to continue with sciences to tertiary level were mainly predicted by interest in science.

The findings of this study were also found to differ from the previous findings for other subject domains. For example, interest value was not found to have a significant direct influence on sustaining the outcomes variables of SEMP. Nevertheless, it has been theoretically identified as a strong predictor of engagement (Seiler 2005) and enrolment plans in various subjects (Barnes 1999; Krapp 2000; Kelly 1988; Levy 2003; Murphy et al. 2006; Murphy and Whitelegg 2006; Wigfield and Eccles 1989).

An unexpected finding in relation to sustaining enrolment plans in relation to physics was that the utilitarian value did not take the leading role. This finding was unexpected since most research on initial enrolment intention shows utility value as the strongest predictor for future decisions to enrol in physics. Instead, this study found that despite the near or average values for utility (see Table 1), performance perceptions was the strongest predictor in sustaining further enrolment plans in physics across the four modules. These results suggest that, while the utility value of physics is the largest determining influence on initial enrolment plans in relation to the subject, it may get superseded by an expectancy variable once students have started studying physics and when they are making decisions about continuing their enrolment in the subject.

A possible reason for this major departure from the previous findings might be that students who had commenced studying physics were already convinced about its value and strategic utility at the time they made their initial choice to study the subject. These students are now trying to decide whether to continue with a second year of physics and prepare for high-stakes examinations in this subject at the end of year 12. This will only make sense if they believe that they are good at the subject. For them, both their present and their future perceptions of self function as a frame of reference. In contrast, students who are making initial enrolment plans are not sure how good they might be in the subject, so other constructs such as the utility value of physics are more integral to their decision-making process. They might have asked themselves “why would I do this subject?” (Eccles 2008) and the value of physics for their future study and career plans, which was evident for them, might not have undergone further change once they started studying the subject in year 11. However, once they started to study year 11 Physics, they appear to have paid more attention to the question “how well can I do the subject?” in order to decide whether to continue with Physics (Eccles 2008), although they still attach high utility value to the subject.

The perceived nature of the subject and characteristics of the students could be other factors. Australian senior secondary physics students are generally high academic achievers with high career and study ambitions (Fullarton and Ainley 2000). Therefore, they might be placing more emphasis on performance perceptions in a given subject in their year 11 curriculum than on any other motivational variable. With physics being perceived as a “hard” subject, this might be being exhibited more obviously. In NSW, senior secondary courses are perceived as career preparation courses (Barnes 1999) and considerable importance is placed on the Australian Tertiary Admission Rank (ATAR) which is calculated at the end of the final year of senior secondary school examinations. There is a trend among students to discontinue the subjects in which they are not achieving well enough to get a high ATAR. This may be another reason for placing high significance on this particular variable.

The negative and significant paths from sexstereo to engage across the majority of the modules (except for waves) suggest that possession of stereotypical attitudes could reduce students’ engagement with physics significantly. Previous research from related studies supports this observation. Barres (2006) argued that the foremost factor to blame for women’s slow advance in science was based on social stereotypes that suggested that women are innately less able than men. This is supported by Spelke’s (2005) research, which suggests that the cognitive abilities of males and females in relation to science are, on average, equal. Similarly, Kelly (1988) observed that cultural influences from society and schools could magnify the concept that science is a “masculine” subject and that this attitude could deter females from participation in science, particularly in physics. The SEMPs developed for this study support the conclusion that a stereotyped attitude that defines physics as a “male domain” can significantly reduce student engagement with the enacted physics curriculum.

Across all modules, interest value was not identified as a significant direct influence on outcome variables. In Woods’ (2008) study of physics enrolment, interest value was not identified as a direct significant factor, indicating that interest value did not directly influence the initial physics enrolment motivation of females. Rather, its effects were indirect, via mediating variables. Though Woods’ study was confined to female students, this study supports her conclusions.

Some of the results of this investigation partially supported existing research findings. Barnes’ (1999) enrolment model for physics identified performance expectation as one of the significant predictors of initial enrolment intentions in physics. However, it was not the largest predictor of physics enrolment plans in Barnes’ model. Cleaves (2005) found that even higher achieving 13–16 year olds underestimated their own science ability. According to Cleaves, students’ confidence in their own ability to do science is a major factor other than reasons of interest and enjoyment in relation to enrolment plans. In support of this, Hipkins and Bolstad (2005) argued that “deflated self-esteem with respect to science achievement is one of the powerful factors that mitigated against deciding to continue in science” (p. 35). Likewise, Woods (2008) found that perceived high self-concept of ability positively predicted female initial physics enrolment, though utility value was a stronger determining factor. She observed that girls with belief in their ability to successfully meet the requirements of senior physics were more likely to enrol in physics than girls who believed physics would be too difficult (see also Gill and Bell 2013; Mujtaba and Reiss 2013). Thus, self-concept of ability in science has been identified as one of the significant influences on science enrolment intentions. In the present study, self-concept of ability and task difficulty were subsumed by the performance perceptions construct. Hence, by acknowledging the significance of performance perceptions, the present study supports the conclusion from related studies.

The most significant contribution of the investigation to the enrolment research is that it advances the EV research knowledge that, while values may be significant, it is the expectancies that largely predict students’ sustained choice intentions in physics at a topic-specific level. The module specificity of SEMP allowed for the mapping of the motivation patterns more sensitively, suggesting that the relative influence of motivational precursors vary by topic. This greater sensitivity in measurement is another unique contribution of this investigation to the research on EV theory and sustained physics enrolment intentions.

Implications for Educational Practitioners

The theoretical and empirical viabilities of SEMP contribute to physics teachers’ pedagogical knowledge. Practitioners need to be aware of the motivational properties of students’ expectancies in physics (perfperc) in determining sustained engagement and subsequent sustained enrolment plans in physics. The school and classroom environments are vital contexts that can enhance the performance perceptions of physics students. Educators can improve their classroom practices by utilising pedagogical approaches that enhance students’ performance perceptions in physics classrooms as we now know this can positively influence the students’ sustained engagement and intentions to enrol in physics.

The results of this study suggest that in physics classrooms it is important that students feel competent and achieve success in physics. The fears of poor performance arise in part from the competitive mindset generated by the Australian Tertiary Admissions Rank (ATAR) and the narrow view of success that it encourages. While teachers cannot alter the facts of ATAR and end-of-year summative assessment, they can reduce the anxieties which that mindset tends to produce. Teachers need to consider teaching and assessment practices that boost students’ performance perceptions. For example, teachers could conceptualise success in alternative ways rather than simply emphasising high achievement in summative assessment tasks. If students are in a classroom where success is defined in terms of self-improvement rather than performing relative to peers then all students have the chance of feeling success and potentially altering their performance perceptions. Physics teachers should encourage students to take on difficult tasks and make mistakes and teach the students to understand this as part of the learning process. This is an important motivational lesson since adolescents, especially high-achieving girls, tend to endorse more of a fixed view of ability than their male counterparts, in science (Dweck 2007). Students those who think of ability as a fixed trait are at greater risk of negative academic outcomes such as decreased confidence and performance impairment (Good et al. 2003). Consequently, there is utility in senior secondary physics teachers demonstrating to their students that effort and outcome co-vary. Success in physics can be acquired through hard work and perseverance. Other recommendations for practitioners would be to encourage cooperative and collaborative learning activities. These may promote students to work together to solve tasks rather than to compete against each other. Social interactions may provide more opportunities for students to succeed and to share the feeling of success. These positive feelings may have a flow on effect and lead students to be more enthusiastic about the subject.

Another suggestion would be for practitioners to be aware of the sex stereotypes that appear to significantly reduce student engagement and participation. Therefore, learning experiences and teaching practices that discourage the development of such attitudes should be incorporated into physics instruction. In addition teachers need to be cautious about making false evaluations that lead to gender differentiated expectations and classroom practices. Physics classroom observations indicate that teachers are practising such expectations although they are unaware of them. For example, paying more attention to contributions from any particular gender, setting higher expectations for any particular gender’s achievements than the others (Hoffmann 2002; Millar and Toscano 2006), attributing each gender’s success to different reasons (Spear 1984, 1985, cited in Hoffman 2002 p. 454), and encouraging students in a gender stereotyped manner (Whyte 1986, cited in Hoffmann 2002) could enhance sex stereotyped attitudes in physics classes. Participants in this study showed lower levels of such gender-biased attitudes towards all four physics modules (Table 1). That is, students do not perceive physics to be more suitable for one gender over the other, and therefore it is important that teachers hear this message and employ appropriate strategies that favor both genders instead of reproducing gender-stereotyped practices.


The key findings of this study are the theoretical and empirical viability of the SEMPs. The SEMPs indicate that it is the expectancies of success that offers a strong explanation for sustained engagement and enrolment in relation to a “demanding” subject like physics in the senior secondary years. This finding applied across all four physics modules substantiating the significance of students’ perceptions of their performance. The viability of this model could be tested with other science subjects to determine its generalisability. A fine-grained analysis at a topic-specific level revealed a new pattern or relationships. Perhaps conducting research at a more specific level within a science subject could bring out the true effects of motivation and engagement, which might get masked at a subject or domain general level. The variation in the relative influences of the predictor variables of the four SEMPs tested in this study indicates that this might in fact be the case. Although the low representation of females in the sample was expected, it did limit the ability to examine the SEMPs across males and females. The female sample size was smaller than 100, which is the minimum required for a separate Structural Equation Modelling analysis. Hence, a comparative analysis of the motivational patterns of male and female students could not be conducted. Future studies that involve a larger sample could investigate whether there are gender differences for the SEMP.


  1. Abraham, J., & Barker, K. (In press). Sustaining young people’s enrolment intentions in relation to physics: Development and validation of a tool. The Australian Journal of Educational & Developmental Psychology.Google Scholar
  2. Ainley, M. (2004). What Do We Know about Student Participation and Engagement? Melbourne: Paper presented at the Australian Association for Research in Education.Google Scholar
  3. Ainley, J., Kos, J., & Nicholas, M. (2008). Participation in Science, Mathematics and Technology in Australian Education (Research Monograph No 63). Camberwell, Victoria, Australia: The Australian Council for Educational Research. Scholar
  4. Barnes, G. R. (1999). A Motivational Model of Enrolment Intentions in Senior Secondary Science Courses in New South Wales Schools. Doctoral dissertation, University of Western Sydney, Macarthur.Google Scholar
  5. Barres, B. A. (2006). Does gender matter? Nature, 442(7099), 133–136. doi:10.1038/442133a.CrossRefGoogle Scholar
  6. Bennett, J., Lubben, F., & Hampden-Thompson, G. (2013). Schools that make a difference to post-compulsory uptake of physical science subjects: Some comparative case studies in England. International Journal of Science Education, 35(4), 663–689. doi:10.1080/09500693.2011.641131.CrossRefGoogle Scholar
  7. Bultitude, K., Grant, L., Daly, A., & Thomson, C. (2010). Girls into physics (invited presentation) diversity in STEM conference. London, UK: Institute of Physics.Google Scholar
  8. Byrne, B. M. (1998). Structural equation modelling with LISREL, PRELIS, and SIMPLIS: Basic concepts, applications and programming. Mahwah, NJ: Erlbaum.Google Scholar
  9. Cleaves, A. (2005). The formation of science choices in secondary school. International Journal of Science Education, 27(4), 471–486. doi:10.1080/0950069042000323746.CrossRefGoogle Scholar
  10. Dweck, C. S. (2007). Is math a gift? Beliefs that put females at risk. In S. J. Ceci & W. M. Williams (Eds.), Why aren't more women in science: Top researchers debate the evidence (pp. 47–55). Washington, DC, US: American Psychological Association.CrossRefGoogle Scholar
  11. Eccles, J. S. (1984). Sex differences in mathematics participation (Vol. 2). Greenwich, CT: JAI Press Inc.Google Scholar
  12. Eccles, J. S. (1987). Gender roles and Women's achievement-related decisions. Psychology of Women Quarterly, 11(2), 135–172. doi:10.1111/j.1471-6402.1987.tb00781.x.CrossRefGoogle Scholar
  13. Eccles, J. (1989). Gender and thought: Psychological perspectives. In M. Crawford & M. Gentry (Eds.), Bringing young women into math and science. Berlin, Heidelberg, New York: Springer.CrossRefGoogle Scholar
  14. Eccles, J. S. (2008). Can Middle School Reform Increase High School Graduation Rates?: California Dropout Research Project Report #12.Google Scholar
  15. Eccles, J. S., & Wigfield, A. (1995). In the mind of the actor: the structure of adolescents' achievement task values and expectancy-related beliefs. Personality and Social Psychology Bulletin, 21(3), 215–225. doi:10.1177/0146167295213003.CrossRefGoogle Scholar
  16. Eccles, J., Adler, T. F., Futterman, R., Goff, S. B., Kaczala, C. M., Meece, J. L., & Midgely, C. (1983). Expectancies, values and academic behaviors. In J. T. Spence (Ed.), Achievement and achievement motivation (pp. 75–146). San Francisco, CA: Freeman.Google Scholar
  17. Eccles, J., Barber, B., Updegraff, K., & O'Brien, K. (1998). An Expectancy-Value Model of Achievement Choices: The Role Of Ability Self-Concepts, Perceived Task Utility, and Interest in Predicting Activity Choice and Course Enrolment. In A. Krapp, A. Renninger & J. Baumert (Eds.), Interest and learning: Proceedings of the Seeon Conference on Interest and Learning (pp. 267–279): Institute for Science Education at the University of Kiel (IPN).Google Scholar
  18. Eccles, J., O’Neill, S., & Wigfield, A. (2005). Ability Self-Perceptions and Subjective Task Values in Adolescents and Children. In K. A. Moore & L. H. Lippman (Eds.), What do children need to flourish? (Vol. 3, pp. 237–249). US: Springer.CrossRefGoogle Scholar
  19. Feather, N. T. (1988). Values, valences, and course enrolment: Testing the role of personal values within an expectancy-valence framework. Journal of Educational Psychology, 80(3), 381–391.CrossRefGoogle Scholar
  20. Fullarton, S., & Ainley, J. (2000). Subject choice by students in year 12 in Australian secondary schools. LSAY research reports. Longitudinal surveys of Australian youth research report. Camberwell, Victoria: Australian Council for Educational Research.Google Scholar
  21. Fullarton, S., Walker, M., Ainley, J., & Hillman, K. (2003). Longitudinal studies of Australian youth: research report 33: Patterns of participation at year 12. Camberwell, Victoria: Melbourne: Australian Council for Educational Research.Google Scholar
  22. Garg, K., & Gupta, B. (2003). Decline in science education in India: a case study at +2 and undergraduate level. Current Science, 84(9), 1198–1201.Google Scholar
  23. Gill, T., & Bell, J. F. (2013). What factors determine the uptake of a-level physics? International Journal of Science Education, 35(5), 753–772. doi:10.1080/09500693.2011.577843.CrossRefGoogle Scholar
  24. Good, C., Aronson, J., & Inzlicht, M. (2003). Improving adolescents' standardized test performance: an intervention to reduce the effects of stereotype threat. Journal of Applied Developmental Psychology, 24(6), 645–662. doi:10.1016/j.appdev.2003.09.002.CrossRefGoogle Scholar
  25. Goodrum, D., Hackling, M., & Rennie, L. (2001). The status and quality of teaching and learning of science in Australian schools. Canberra: Department of Education, Training and Youth Affairs.Google Scholar
  26. Goto, M. (2001). ‘Japan’, in International Bureau for Education (IBE). Geneva: IBE, UNESCO.Google Scholar
  27. Hills, A. M. (2008). Foolproof guide to statistics using SPSS. Frenchs Forest, Australia: Pearson Education.Google Scholar
  28. Hipkins, R., & Bolstad, R. (2005). Staying In Science: Students’ Participation In Secondary Education and on Transition to Tertiary Studies Wellington: New Zealand Council For Educational Research.Google Scholar
  29. Hipkins, R., Roberts, J., Bolstad, R., & Ferral, H. (2006). Staying in science 2: Transition to tertiary study from the perspectives of New Zealand year 13 science students. Report prepared for the New Zealand ministry of research science and technology. Wellington: MoRST.Google Scholar
  30. Hoffmann, L. (2002). Promoting Girls' interest and achievement in physics classes for beginners. Learning and Instruction, 12(4), 447–465. doi:10.1016/s0959-4752(01)00010-x.CrossRefGoogle Scholar
  31. Jöreskog, K. G., & Sörbom, D. (1996). LISREL 8: Structural equation modelling with the SIMPLIS command language. Chicago: Scientific Software International.Google Scholar
  32. Kelly, A. (1988). Option choice for girls and boys. Research in Science and Technological Education, 6(1), 5–23.CrossRefGoogle Scholar
  33. Khalijah, M. S. (2004). Role of physics community for the development and advancement of physics education in the globalization Era. Indonesian Journal of Physics, 15(1), 7–13.Google Scholar
  34. Kline, R. B. (1998). Principles and practice of structural equation modelling. New York: The Guilford Press.Google Scholar
  35. Krapp, A. (Ed.). (2000). Interest and human development during adolescence: an educational–psychological approach. London: Elsevier.Google Scholar
  36. Krogh, B. L., & Thomsen, P. V. (2005). Studying students’ attitudes towards science from a cultural perspective but with a quantitative methodology: Border crossing into the physics classroom. International Journal of Science Education, 27(3), 281–302. doi:10.1080/09500690412331314469.CrossRefGoogle Scholar
  37. Levy, I. (2003). 10th Graders' preferences for science specialization in terms of their attitude to the subject. MA: Tel Aviv University.Google Scholar
  38. Lyons, T. (2006). The puzzle of falling enrolments in physics and chemistry courses: Putting some pieces together. Research in Science Education, 36(3), 285–311. doi:10.1007/s11165-005-9008-z.CrossRefGoogle Scholar
  39. Lyons, T., & Quinn, F. (2010). Choosing Science: Understanding the Declines in Senior High School Science Enrolments. Research Report to the Australian Science Teachers Association: UNE.Google Scholar
  40. Meece, J., Wigfield, A., & Eccles, J. (1990). Predictors of math anxiety and its influence on young Adolescents' course enrolment intentions and performance in mathematics. Journal of Educational Psychology, 82(1), 60–70. citeulike-article-id:1623479.CrossRefGoogle Scholar
  41. Millar, V., & Toscano, M. (2006). Girls in Physics. Australian School Innovation in Science,Technology and Mathematics (ASISTM) Project.: Retrieved from : stav2006/A2.ppt
  42. Mujtaba, T., & Reiss, M. J. (2013). What sort of girl wants to study physics after the age of 16? Findings from a large-scale UK survey. International Journal of Science Education, 35(17), 2979–2998. doi:10.1080/09500693.2012.681076.CrossRefGoogle Scholar
  43. Murphy, P., & Whitelegg, E. (2006). Girls and physics: Continuing barriers to ‘Belonging’. Curriculum Journal, 17(3), 281–305.CrossRefGoogle Scholar
  44. Murphy, P., Lunn, S., & Jones, H. (2006). The impact of authentic learning on students' engagement with physics. Curriculum Journal, 17(3), 229–246.CrossRefGoogle Scholar
  45. Osborne, J., Simon, S., & Collins, S. (2003). Attitudes towards science: a review of the literature and its implications. International Journal of Science Education, 25(9), 1049–1079.CrossRefGoogle Scholar
  46. Rennie, L., & Parker, L. (1996). Placing physics problems in real-life context: Students' reactions and performance. Australian Science Teachers Journal, 42(1), 55–59.Google Scholar
  47. Russell, V. J., Ainley, M., & Frydenberg, E. (n. d.). Schooling Issues Digest: Student Motivation and Engagemen. Retrieved from
  48. Russell, J., Mackay, A., & Jane, G. (2003). Messages from the Myrad: Improving the middle years of schooling. Jolimont, Melbourne: IARTV.Google Scholar
  49. Schafer, J. L., & Graham, J. W. (2002). Missing data: Our view of the state of the art. Psychological Methods, 7(2), 147–177.CrossRefGoogle Scholar
  50. Science, Technology, Engineering and Math: Education for Global Leadership. (n. d.). Retrieved May 15, 2014, from
  51. Seiler, G. (2005). All My Life I Been Po’: Oral Fluency As A Resource for Science Teaching and Learning. In K. Tobin, R. Elmesky, & G. Seiler (Eds.), Improving urban science education: New roles for teachers, students and researchers (pp. 141–158). New York: Rowman and Littlefield.Google Scholar
  52. Spall, K., Barrett, S., Stanisstreet, M., Dickson, D., & Boyes, E. (2003). Undergraduates' views about biology and physics. Research in Science and Technological Education, 21(2), 193–208.CrossRefGoogle Scholar
  53. Spelke, E. S. (2005). Sex differences in intrinsic aptitude for mathematics and science? a critical review. American Psychologist, 60(9), 950–958. doi:10.1037/0003-066x.60.9.950.CrossRefGoogle Scholar
  54. Stokking, K. M. (2000). Predicting the choice of physics in secondary education. International Journal of Science Education, 22(12), 1261–1283.CrossRefGoogle Scholar
  55. Wigfield, A., & Eccles, J. S. (1989). Relations of Expectancies and Values to Students' Math Grades and Intentions. Paper presented at the Paper presented al the Meeting of the American Educational Research Association, San Francisco.Google Scholar
  56. Wigfield, A., & Eccles, J. S. (1992). The development of achievement task values: a theoretical analysis. Developmental Review, 12(3), 265–310. doi:10.1016/0273-2297(92)90011-p.CrossRefGoogle Scholar
  57. Wigfield, A., & Eccles, J. S. (2000). Expectancy-value theory of achievement motivation. Contemporary Educational Psychology, 25(1), 68–81. doi:10.1006/ceps.1999.1015.CrossRefGoogle Scholar
  58. Willms, J. D., (2003). Student Engagement at School: A Sense Of Belonging And Participation: Results from PISA 2000: OECDGoogle Scholar
  59. Woods, C. (2008). Factors Predicting High Ability Girls’ Enrolment in Year 11 and 12 Physics and Intermediate Mathematics in Regional and Rural North Queensland. Paper presented at the AARE Conference, Brisbane.Google Scholar

Copyright information

© Springer Science+Business Media Dordrecht 2014

Authors and Affiliations

  1. 1.School of EducationUniversity of Western SydneySydneyAustralia

Personalised recommendations