Motivation is a crucial condition for learning. Motivation is related to among others school performance (Froiland & Worrel, 2016; Gottfried et al., 2008; Scales et al., 2020; Taylor et al., 2014), engagement (Froiland & Worrel, 2016), dropout and self-concept at school and in general (Gottfried et al., 2008). However, student motivation declines during adolescence (Gnambs & Hanfstingl, 2016) and tend to be lower in secondary education compared to elementary education (Corpus et al., 2009). The intrinsic motivation of students with special educational needs (SEN) may be even more at risk (Shogren et al., 2019; Zisimopoulos & Galanaki, 2009). Therefore, an important question is what adolescents, in particular students with SEN, need to feel intrinsically motivated at school.

Most of the research in educational psychology takes a between-person approach towards such questions, focusing on inter-individual differences. A within-person approach, in contrast, measures concepts several times for the same person, to explore intra-individual development such as relations between variables within individuals (Murayama et al., 2017). Only if very strict assumptions are met, generalization of results from the group level to the individual level is possible. This is known as the concept of ergodicity (Molenaar, 2004).

Last years, the limitations of a between-person approach are more and more recognized in psychology. Although between-person studies revealed interesting findings in educational psychology, these studies offer limited insight in intra-individual psychological processes (Murayama et al., 2017). Researchers thus conclude that a within-person approach is needed to draw conclusions on individual psychological processes (Hamaker, 2012; Molenaar, 2004). This means that variables, situations or time-points within individuals are the units of analysis, instead of the individuals themselves. Recent studies with intense intra-individual data show that students have different motivational profiles (eg. Gillet et al., 2017; Lee et al., 2020; Ramos et al., 2021; Wormington et al., 2012), indicating the importance of this kind of research.. Therefore, this study will, alongside a between-person perspective, take a within-person approach to motivation. In particular we will focus on the self-determination theory (SDT), as attention for individual processes is also needed in SDT-research (Ryan & Deci, 2017). The SDT also describes individual psychological processes, such as the relationship between basic psychological need satisfaction and intrinsic motivation. Taking a within-person approach, we gain more knowledge on the extent to which the relevance of basic psychological needs for motivation is different for different individuals. Furthermore, we explore the presence or absence of SEN to further clarify possible individual differences.

The SDT (Deci et al., 1991; Ryan & Deci, 2000, 2017) is one of the leading theories of motivation. A central concept in the SDT is intrinsic motivation, which refers to ‘activities that are done for their inherent satisfactions’ (Ryan & Deci, 2017, p. 117). Intrinsic motivation is the most self-determined type of motivation on the self-determination continuum (Deci & Ryan, 2002). In contrast, extrinsic motivation holds behaviour that is instrumental towards a certain outcome and is less autonomous. The more an activity is internalized, the more self-determined the behaviour. This process of internalization is facilitated by support of the basic psychological needs (Deci & Ryan, 2002). To address this process, we investigate the relationship between basic psychological need satisfaction and intrinsic motivation, as the most self-determined type of motivation.

The SDT states that individuals have three basic psychological needs: the needs for autonomy, competence and relatedness (Ryan & Deci, 2017). According to the SDT, individuals need ownership of the things they do (autonomy), the feeling that they can achieve their goals (competence) and significant relationships with people around them (relatedness) (Ryan & Deci, 2000). Fulfilment of the three needs is positively related to intrinsic motivation. SDT states that the basic psychological needs are ‘etic universal’: ‘characteristics or processes that can be empirically identified as cross-culturally valid’ (Ryan & Deci, 2017, p. 566). Moreover, fulfilling the needs will have positive effects on well-being, regardless of how an individual values or desires the specific need (Chen et al., 2015).

The SDT is applicable in many contexts, including educational settings. Support of the basic psychological needs leads to positive student outcomes in education. Need support at school enhances students’ learning and school performance (Ryan & Deci, 2017), and situational interest (Minnaert et al., 2011). When students experience that their needs are being met, they are more likely to feel motivated for school and to show engagement in the classroom (Stroet et al., 2013). Need satisfaction and motivation seem to go together: a decline in academic motivation between ages 11 and 16 is associated with a decline in the satisfaction of the needs for autonomy, competence and relatedness (Gnambs & Hanfstingl, 2016).

The last years, the SDT-perspective was also applied to research on students with SEN: students who are in need for extra support in the regular classroom. In the Netherlands, as in most countries, no nationwide criteria are used to indicate students’ SEN. Instead, schools focus on what each individual student needs. Based on teachers’ indications, approximately a quarter of the students in regular secondary education are perceived as students with SEN, mostly showing internalizing and externalizing behavioural problems, and problems in work engagement (Smeets et al., 2019).

Adolescents with SEN seem to be more motivated by external factors compared to internal factors (Shogren et al., 2019). Furthermore, students with learning disabilities have lower levels of intrinsic motivation than their classmates (Zisimopoulos & Galanaki, 2009). Given the importance of intrinsic motivation, it is thus important to investigate were this lower intrinsic motivation comes from.

Next to motivation, there are doubts about the fulfilment of the basic psychological needs of students with SEN in regular education. For example, children with Attention Deficit Hyperactivity Disorder (ADHD) in regular primary education report their classroom as more controlling than their classmates (Rogers & Tannock, 2018). Besides, children with ADHD (Rogers & Tannock, 2018) and learning disabilities (Zisimopoulos & Galanaki, 2009) in primary education feel in general less competent at school than their classmates. To the contrary, students with SEN may be more cognitively challenged in regular education compared to special education, which may have a positive effect on their feelings of competence (Myklebust, 2007). Lastly students with ADHD feel less related to their teachers compared to their classmates (Rogers & Tannock, 2018) and students with behavioural problems or intellectual disabilities are at risk of social exclusion by their classmates (de Boer et al., 2012). However, other studies show that students with SEN do have positive relationships with peers as well as with adults at school (López et al., 2016; Rose & Shevlin, 2017).

According to the SDT, the basic psychological needs for autonomy, competence and relatedness are universal, which implies that students with SEN have the same basic psychological needs for autonomy, competence and relatedness as their classmates (Ryan & Deci, 2017). For example, for people with mild intellectual disabilities, fulfilment of these needs is positively related to seeing the usefulness of a task, and stimulating autonomy leads to even higher task value (Emond Pelletier & Joussemet, 2017). Other research shows that feelings of competence and autonomy seem beneficial for adjustment in school for both students with learning disabilities and those with emotional disabilities (Deci et al., 1992). However, although satisfaction of all the three needs is related to self-determination for adolescents with ADHD, autism and/or learning disabilities, only satisfaction of the need for competence is correlated with intrinsic motivation (Shogren et al., 2019). There may also be some differences between subgroups of students, as autonomy seems to be more important for students with emotional disabilities and competence for students with learning disabilities (Deci et al., 1992).

Given the importance of intrinsic motivation for learning, the importance of basic psychological need satisfaction for intrinsic motivation, and the doubts about the intrinsic motivation of adolescents, especially students with SEN, it is important to study the relationship between basic psychological need satisfaction and intrinsic motivation of typically developing students and students with SEN in secondary education. The SDT, like many theories in educational psychology, describes within-person relationships of psychological processes: satisfaction of the basic psychological needs leads to intrinsic motivation. This means that a within-person approach is needed to study this relationship (Murayama et al., 2017). Besides, the difficulties that teachers have with adapting to the needs of a diverse student population (van Grinsven & van der Woud, 2016) show the importance of looking at the relationship between need satisfaction and intrinsic motivation at an individual level. Therefore, alongside a between-person approach, this study takes a within-person approach to the SDT.

The first goal of the study is to investigate if the relationship at the group level also holds at the level of the individual student. The second goal is to investigate the role of SEN in within-person differences. Therefore, the following research questions will be addressed:

  • To what extent are basic psychological need satisfaction and intrinsic motivation related at the group level?

  • Are there individual differences in the relationship between basic psychological need satisfaction and intrinsic motivation?

  • Does the relationship between basic psychological need satisfaction and intrinsic motivation differ for students with and without SEN?

It is hypothesized that the largest amount of variability in basic psychological need satisfaction and intrinsic motivation comes from within individual differences. Besides, a positive association between need satisfaction and intrinsic motivation is expected, with individual differences in the strength of this relationship. Students SEN might play a role in these individual differences.

Method

Design

To take a within-person approach, intrinsic motivation and basic psychological need satisfaction of the students were measured with the experience sampling method (ESM). ESM is ‘a research procedure for studying what people do, feel and think during their daily lives’ (Larson & Csikszentmihalyi, 2014, p. 21), which can provide unique data compared to general questionnaires (Suveg et al., 2010). With ESM, it is possible to measure psychological processes in the daily context, near the time they occur, and to study patterns over time (Iida et al., 2012). Measuring experiences over time allows to describe patterns at an individual and at the group level (Csikszentmihalyi & Larson, 1987). ESM thus allows to take a within-person approach to the relationship between basic psychological need satisfaction and intrinsic motivation. In our study, this means that students repeatedly completed a diary questionnaire measuring intrinsic motivation and need satisfaction (see the ‘Procedure’ section).

Participants

The present study is part of the research project Differentiation Inside Out. The students belonged to eleven classes from seven schools for regular secondary education spread across the Netherlands. In the current study, 215 students from nine classes participated, because the teachers of two classes, and as such the students, dropped out. The students attended the second year of pre-vocational secondary education (8th grade) in the 2018–2019 school year. Their track provides general education preparing for further vocational education and a quarter of all students in secondary education follow this same track (Inspectie van het Onderwijs, 2020). In the second year, students already know the school and their class but do not have chosen a specific orientation. Students with SEN are over-represented in this track, as compared to the higher tracks in secondary education (Inspectie van het Onderwijs, 2020; Smeets et al., 2019), which makes this track especially interesting.

In our sample, 27% of the students had at least one special educational need, which is equivalent to earlier research (Smeets et al., 2019). The most prevalent needs were learning problems, like dyslexia. The distribution of boys and girls in the sample was almost even (56% girls). The majority of the students participated with their Dutch teacher, one class with their mathematics teacher. In classes of other subjects, they did not participate in the research, to make sure that their reported motivation did not vary because of the subject. Dutch and mathematics were chosen as we suppose that lessons in mother language and mathematics are comparable outside of the Netherlands.

The research was divided in three waves: the first wave in November and December, the second in January and February, and the third in March and April. Students were only included in the analysis if they completed at least one diary questionnaire in each wave of the study. Sixty-nine students were excluded based on this criterion, meaning that the final sample of this study consisted of 146 students (56% girls, 25% students with SEN). An overview of the selection process of the sample is given in Fig. 1. Characteristics of the samples are shown in Table 1.

Fig. 1
figure 1

Flow chart sample current study

Full size image
Table 1 Characteristics of the sample
Full size table

To check if the included sample was representative, we performed missing data analysis on the two groups: the students who did complete a diary in every wave (N = 146) and the students who did not complete a diary in every wave (N = 69). There seemed to be some class differences (F = 11.70, p < 0.001), meaning that students of some classes were less likely drop-out than students of other classes. This might be explained by active participation and stimulation of the teachers to complete the questionnaire. For example, some teachers stimulated the students to complete the questionnaire at the end of class, which is positive for the response rate of the students. Boys and girls as well as students with and without SEN were equally represented in both groups (χ2 < 0.01, p = 0.961; χ2 = 0.80, p = 0.370). Besides, we calculated mean scores of each student on intrinsic motivation and basic psychological need satisfaction. These did not significantly differ between the groups on intrinsic motivation (t =  − 1.74, p = 0.084), autonomy (t =  − 1.45, p = 0.151), competence (t =  − 1.59, p = 0.114), and relatedness with classmates (t =  − 0.15, p = 0.885). Besides, effect sizes were (very) small: Cohen’s d is 0.26 for intrinsic motivation and competence, 0.23 for autonomy and 0.02 for relatedness with classmates. Included students did show significantly higher scores on relatedness with the teacher (t =  − 2.22, p = 0.029). However, this effect size was still small (Cohen’s d = 0.39).

Procedure

Schools and teachers were recruited via social media and contacts of the research team. Inclusion criteria for both teachers and students were used: teachers needed to teach Dutch, English or mathematics and their students needed to attend the second year of pre-vocational secondary education. Teachers and parents of the students gave permission for participation in the study. The Ethics Committee of Pedagogical & Educational Sciences of the University of Groningen approved the project (8th of October 2018).

To investigate individual differences in the relationship between intrinsic motivation and basic psychological need satisfaction, we used ESM. Students repeatedly completed a diary questionnaire about their experiences of intrinsic motivation and need satisfaction during a specific lesson. The data collection lasted for approximately 20 weeks, roughly from the end of October (roughly 2 months after the start of the school year) until the end of April (roughly 2 months before the end of the school year). Duration varied between 18 and 23 weeks of data collection at school, depending on the exact start and end date of the project in each class. Students completed the same short questionnaire at the end of every lesson of the participating teacher. The questions specifically addressed the students’ motivation in this lesson. The number of measurement points each week differed between classes, as lessons were scheduled two or three times a week. Questionnaires were sent 15 min before the end of the lesson by email or text message or both, depending on the participants, and were completed online. In most of the classes, the teachers allowed the students to complete the questionnaire in the last 5 min of the lesson to prevent missing data and memory loss.

Instruments and variables

Special educational needs

In the Netherlands, as in many countries, no nationwide criteria are used to indicate the special educational needs of students. Instead of a medical model based on official diagnoses, schools focus on what an individual student needs. Due to the absence of a system for determining SEN, there are no clear cut-off criteria. Consequently, it is hard to determine the exact number of students with SEN in a regular classroom (Inspectie van het Onderwijs, 2020; Ledoux et al., 2020). Teachers’ indications of students’ SEN are often used to assess the special educational needs of students (Smeets et al., 2019). In this research, this method is also adhered. Teachers were asked to provide a list indicating which students in their class had SEN (with or without formal assessment), and to include a brief characterization of which type of SEN. In most cases, the lists of these students were provided by the teachers, with the lists for two classes being provided by the school. As we asked for an open characterization of SEN, the researchers then classified the reported SEN types into assessment categories: learning problems, behavioural problems, and other problems such as physical impairments.

Intrinsic motivation and basic psychological need satisfaction

Intrinsic motivation and basic psychological need satisfaction of the students were measured using the experience sampling method. Students completed a diary questionnaire after each lesson of their teacher in Dutch or mathematics, as explained in the previous section (see the ‘Procedure’ section). The diary questionnaire consisted of 22 items, and every item was a statement about the student’s experiences of the lesson. Answers ranged from 0, meaning strongly disagree, to 100, meaning strongly agree. Two scales of the questionnaire (15 items in total) were used in this study. An overview can be found in Table 2.

Table 2 Diary questionnaire
Full size table

The scale measuring intrinsic motivation was based on the interest/enjoyment scale of the Intrinsic Motivation Inventory. This part of the questionnaire consisted of seven questions. The internal consistency (Cronbach’s alpha) of this scale in the current study is 0.83.

The scale measuring the satisfaction of students’ basic psychological needs was based on the satisfaction scale of the diary version of the Basic Psychological Need Satisfaction and Frustration Scale (Chen et al., 2015; van der Kaap-Deeder et al., 2017). This part of the questionnaire consisted of eight questions. Every basic psychological need was measured by two statements, and a distinction was made between relatedness with the teacher and relatedness with classmates. The internal consistency (Cronbach’s alpha) of this scale in the current study was 0.80 for autonomy, 0.80 for competence, 0.87 for relatedness with the teacher and 0.84 for relatedness with classmates.

Statistical analysis

Before conducting the multilevel analyses that will answer the research questions, we conducted some initial analyses. We did a configural invariance analysis (Byrne, 1998; Milfont & Fischer, 2010) to proof the invariance of intrinsic motivation and basic psychological need satisfaction between the SEN and non-SEN group. To take a within-person perspective to the relationship between basic psychological need satisfaction and intrinsic motivation, we based our analysis on Murayama et al. (2017). First we computed the empty model to investigate variance at each level. To address the first research question (To what extent are basic psychological need satisfaction and intrinsic motivation related at the group level?), a multilevel model was computed with intrinsic motivation as outcome variable, and the basic psychological needs for autonomy, competence, relatedness with the teacher and relatedness with classmates as explanatory variables. In this first model, the slopes were fixed to explore the relationship between satisfaction of the basic psychological needs and intrinsic motivation at the group level. To address the second research question (Are there individual differences in the relationship between basic psychological need satisfaction and intrinsic motivation?), individual differences in the relationship between need satisfaction and intrinsic motivation were explored by investigating random slopes at the student level of the explanatory variables. To answer the third research question (Does the relationship between basic psychological need satisfaction and intrinsic motivation differ for students with and without SEN?), SEN was added to the multilevel model to test if the relationship between need satisfaction and intrinsic motivation differs between students with and without SEN. To conclude, we compare the model fit of all three models comparing the deviances of the models. Doing so, we investigate the importance of individual differences, and the importance of SEN in explaining these individual differences.

Results

Descriptive statistics

The 146 students included in this study (see Fig. 1) completed a total of 4128 diary questionnaires, which means that each student completed on average 28.27 questionnaires. Students without SEN (N = 110) completed on average 28.80 questionnaires (SD = 9.51, minimum = 11, maximum = 56), compared to 26.67 questionnaires (SD = 11.72, minimum = 8, maximum = 56) by students with SEN (N = 36). As explained in the Procedure, not every student could complete as many diaries as other students, as, among others, lessons per week and lessons that were cancelled differed per class.

Initial analyses

Correlation matrices of the items in every measurement wave were computed for students with and without SEN. We included one measurement of each measurement wave per student (N = 438). With a configural invariance analysis (Byrne, 1998; Milfont & Fischer, 2010), we aimed to proof the invariance of intrinsic motivation and basic psychological need satisfaction between the SEN and non-SEN group before doing the multilevel analysis. The analysis revealed that the basic structure of these correlations for both the basic psychological needs and intrinsic motivation were (close to) invariance between the students with and without SEN and across time (all χ2/df < 2.1; CFI’s > 0.96; RMSEA < 0.10; p close fit non-significant in 11 out of 13 invariance analyses). Though, the correlations between the items of basic psychological need satisfaction (Table 3) were somewhat higher among students with SEN, which implies that their needs might be more correlated with each other compared to the needs of students without SEN.

Table 3 Correlations between the basic psychological needs (N = 438)
Full size table

To further investigate the variables of interest, intrinsic motivation and basic psychological need satisfaction, multilevel models with three levels were computed: lessons are nested in students, who are nested in classes. To investigate variance over these levels, empty models were computed, meaning that no explanatory variables are added to the multilevel model. Results regarding intrinsic motivation and basic psychological need satisfaction are depicted in Table 4. The Intraclass Correlations (ICC’s) show that intrinsic motivation and basic psychological need satisfaction mainly fluctuated between students and between lessons. Need satisfaction seemed to be somewhat more variable at the level of the lesson compared to intrinsic motivation, especially the relatedness variables.

Table 4 Empty models of basic psychological need satisfaction
Full size table

To explore the role of SEN in intrinsic motivation and basic psychological need satisfaction, SEN was added to the empty models. Results are shown in Table 5. Students with SEN showed lower intrinsic motivation compared to students without SEN, but this effect of having SEN was not significant (p = 0.128). This also held for the satisfaction of autonomy (p = 0.208), competence (p = 0.159), and relatedness with the teacher and classmates (p = 0.188; p = 0.240).

Table 5 Multilevel models of intrinsic motivation, basic psychological need satisfaction and SEN
Full size table

The relationship between basic psychological need satisfaction and intrinsic motivation

Research question 1: group level

In the first model, the relationship between basic psychological need satisfaction and intrinsic motivation was explored at the group level. Results are shown in Table 6. All needs had a significant positive effect on intrinsic motivation. Competence (β = 0.245; p < 0.001) and relatedness with the teacher (β = 0.241; p < 0.001) seemed to be the most important, followed by autonomy (β = 0.152; p < 0.001) and relatedness with classmates (β = 0.051; p < 0.001).

Table 6 Multilevel models of basic psychological need satisfaction and intrinsic motivation
Full size table

Research question 2: individual differences

In the second model, the relationship between basic psychological need satisfaction and intrinsic motivation was explored at the within-person level. Therefore, the explanatory variables had a random slope at the student level. Results are shown in Table 6.

The main effects of the needs showed the same pattern as in the first model: competence and relatedness with the teacher seemed to have the largest positive effect on intrinsic motivation (β = 0.234, p < 0.001; β = 0.244, p < 0.001) on top of the other needs for autonomy (β = 0.154, p < 0.001) and relatedness with classmates (β = 0.047, p = 0.033).

The random slope model had a better model fit compared to the model with a fixed slope (D = 396.720, p < 0.001). The random slopes were indeed significant, indicating that the relationships between the needs and intrinsic motivation differed between students. The variance seemed to be the largest for competence (s2 = 0.039, p < 0.001), with a + / − 2 SD range from -0.161 to 0.629. The variances in slope for autonomy and relatedness with classmates were smaller, but still significant (s2 = 0.026, p < 0.001; s2 = 0.027, p < 0.001). The slope of autonomy had a + / − 2 SD range from − 0.168 to 0.476, and of relatedness with classmates from − 0.282 to 0.376. The smallest variance was of relatedness with the teacher (s2 = 0.019, p = 0.002), meaning that the + / − 2 SD range was from − 0.032 to 0.520. So looking at the + / − 2 SD range, there were not only differences in the strength of the relationship between need satisfaction and intrinsic motivation, but also the direction of the relationship was different for some students. This especially held for relatedness with classmates and only to a small extent for relatedness with the teacher.

Research question 3: students with SEN

Addressing the third research question, we investigated if having SEN influenced the relationship between basic psychological need satisfaction and intrinsic motivation. In the third model, SEN was added to the first model as an explanatory variable and interaction term with the needs. Results are shown in Table 6.

The main effects of the basic psychological needs showed the same pattern as in the first and second model: competence and relatedness with the teacher had the largest effect on intrinsic motivation (β = 0.225, p < 0.001; β = 0.211, p < 0.001), on top of the other needs for autonomy (β = 0.169, p < 0.001) and relatedness with classmates (β = 0.063, p < 0.001). There was also a significant main effect of SEN (β =  − 6.850, p = 0.045) indicating that students with SEN had lower intrinsic motivation.

The interaction term of SEN and autonomy was negative and significant (β = -0.068, p = 0.023). This implies that the relationship between the satisfaction of the need for autonomy and intrinsic motivation was less strong for students with SEN, but still positive (βBP = 0.169—0.068 = 0.101). To the contrary, the interaction terms of SEN and competence, and SEN and relatedness with the teacher were positive and significant (β = 0.075, p = 0.023; β = 0.096, p = 0.002). The relationships between the satisfaction of the needs for competence and relatedness with the teacher on the one hand, and intrinsic motivation on the other hand were stronger for students with SEN. The interaction term of SEN and relatedness with classmates was not significant, indicating that the relationship between satisfaction of the need for relatedness with classmates and intrinsic motivation did not differ for students with SEN.

To conclude, model 3 had a better model fit compared to model 1 (D = 20.780, p < 0.001), which confirms that having SEN played a role in the relationship between basic psychological need satisfaction and intrinsic motivation. However, model 2 had also a better model fit compared to model 3 (D = 376.010, p < 0.001) and as such had the best model fit. This means that there were individual differences between students in the relationships between the basic psychological needs and intrinsic motivation, and that these differences could only partly be explained by students’ SEN.

Discussion

This research addressed the relationship between basic psychological need satisfaction and intrinsic motivation from both a between-person and a within-person perspective. The first goal of the study was to investigate if the relationship at the group level also holds at the level of the individual student. Attention for individual processes is needed in SDT-research (Ryan & Deci, 2017), but such an approach in still scarce. A within-person perspective with several measurement points per individual allows to analyse relations between variables within individuals (Murayama et al., 2017). Therefore, need satisfaction and intrinsic motivation were measured in real-time and at several lessons with the experience sampling method. In this way, individual differences in the relationship between need satisfaction and intrinsic motivation were investigated. The second goal was to investigate were individual differences in this relationship may come from, exploring the role of SEN. Therefore, students’ SEN was included as a student characteristic that may play a role in explaining individual differences.

The empty models show that variability in intrinsic motivation and basic psychological need satisfaction is the largest at the level of the lesson and the level of the student. The variability of the needs is the largest at the level of the lesson, the variability of intrinsic motivation at the level of the student. This means that need satisfaction is less stable within a person as intrinsic motivation. It may be that students have a base level of motivation, for example based on their general interest in the subject of the lesson, while need satisfaction is more influenced by what happens during a lesson.

In line with the SDT (Ryan & Deci, 2017), at the group level, satisfaction of all basic psychological needs is positively related to intrinsic motivation (research question 1). Competence and relatedness with the teacher have the strongest relationship with intrinsic motivation, followed by autonomy and relatedness with classmates. However, at the within-person level, addressed in the second research question, the relationships between the needs and intrinsic motivation differ between students. The variance is the largest for competence, followed by autonomy and relatedness with classmates, and relatedness with the teacher. Importantly, the relationship does not only differ in strength, but also in direction, meaning that for some students the relationship between need satisfaction and intrinsic motivation is negative. This is contradictory to the claim of the SDT ‘that all individuals have basic needs for competence, relatedness and autonomy’ (Ryan & Deci, 2017, p. 248) and that therefore the effects of need satisfaction are universal, for example regardless of how people value the need (Chen et al., 2015).

The differences in the direction of the relationship between need satisfaction and intrinsic motivation especially hold for relatedness with classmates, meaning that some students may experience a negative influence of satisfaction of this need on their motivation. This is unexpected, as research found that relatedness with friends is important for motivation (Legault et al., 2006), and that students can positively influence each other (Gremmen et al., 2017). For example, high-achieving students can have a positive influence on their friends’ grades. This may also influence low-achievers: it may be that students want to pass the year to stay in the same class, and therefore get more motivated to learn (Gremmen et al., 2017).

Answering the third research question, the relationship between the satisfaction of the need for autonomy and intrinsic motivation is less strong for students with SEN, but still positive. To the contrary, the relationship between the satisfaction of the need for competence and intrinsic motivation is stronger for students with SEN. The relationship between the satisfaction of the need for relatedness with the teacher and intrinsic motivation is also stronger for students with SEN. The relationship between the satisfaction of the need for relatedness with classmates and intrinsic motivation is the same for students with and without SEN. In conclusion, all basic psychological needs seem to have a positive effect on the motivation of students with SEN. However, compared to their classmates, satisfaction of autonomy is less important, while satisfaction of competence and relatedness with the teacher are more important.

Shogren et al. (2019) found that only satisfaction of the need for competence is correlated with intrinsic motivation for adolescents with ADHD, autism and/or learning disabilities. We indeed find the important role of competence for students with SEN, as the relationship between competence and intrinsic motivation was stronger for these students. However, our study does not support the finding that competence is the only need related to intrinsic motivation for students with SEN, as we also find effects of relatedness with the teacher and classmates for these students. The effect of autonomy is smaller compared to students without SEN, but still positive.

To conclude, although all basic psychological needs are positively related to intrinsic motivation at the group level, individual differences in the strength and direction of these relationships exist. This addresses the importance of a within-person perspective: not every need is important for each individual student’s motivation. Students’ SEN play a role, but cannot fully explain these individual differences between students: only differences in strength are found, not in direction. The importance of the needs for intrinsic motivation does differ for students with SEN, with autonomy being less important, and competence and relatedness with the teacher extra important.

Limitations and future research

This study has several limitations and reveals recommendations for future research. First, we only measured need satisfaction, not need frustration. Research on need frustration has grown last years, as it is something different than the absence of need satisfaction (Vansteenkiste & Ryan, 2013): need frustration ‘involves an active threat of the psychological needs’ (Vansteenkiste et al., 2020, p. 9). Need frustration is known to be related to negative outcomes such as ill-being, constricted functioning and externalizing problems (Vansteenkiste & Ryan, 2013). Therefore, it may be that need frustration is related to more extrinsic types of motivation, as was found for students with SEN (Shogren et al., 2019). It would be interesting to also explore the relationship between basic psychological need frustration and intrinsic motivation from a within-person perspective.

Second, we did include lessons at the third level in the multilevel model, as variance in motivation and need satisfaction was explained at this level. However, lessons may not be fully independent of each other: motivation may not only be influenced by what happens in a certain lesson, but also by what happened in the previous lesson. What happened in the lesson 1 day before may be of more influence than what happened in a lesson 3 months ago. Future research may investigate these lesson-to-lesson relationships.

Third, in the Netherlands, the system of care for students with SEN changed and no nationwide criteria are used to decide which students need extra support. Since then, it is hard to determine the amount of students with SEN in a regular classroom (Inspectie van het Ledoux et al., 2020; Onderwijs, 2020). Therefore, it may be that teachers and schools participating in this research differed in the way they interpreted the term ‘students with SEN’. Besides, students with SEN are a heterogeneous group, including students with learning problems, externalizing and internalizing problems, Autism Spectrum Disorder and physical disabilities. Future research may look at the motivation and need satisfaction of students with several SEN in more detail, for example in case studies and/or by focusing on a certain type of disability.

Lastly, although we found high correlations between the basic psychological needs, we deliberately added these variables as separate predictors to the multilevel models. The self-determination theory acknowledges that the needs are correlated, but this does not mean that they are theoretically and empirically the same. The former is grounded by definition (Ryan & Deci, 2017). Concerning the latter, a possible statistical solution for multicollinearity is measuring the needs as one construct. This implies, however, the loss of important information and the acceptance of the false assumption of additivity of the scores on the basic psychological needs. Our results support the theoretical underpinning of the distinct needs by showing a unique contribution of each of the needs to intrinsic motivation. Moreover, the strengths of intercorrelation between the basic psychological needs is lower in the typically developing students (non-SEN group) compared to the SEN group (as shown in Table 3), indicating less common variance among the non-SEN students compared to the SEN students. The latter is explicitly taken into account in the multilevel analysis by adding SEN into the model to test the unique explained variance of the needs.

For further research, it is interesting to focus on low motivated students, as these students might be at risk. What factors influence their lack of motivation? Although we did not find a significant difference in motivation between included and dropped-out students, lower mean scores were found and we might have lost some unmotivated students. Instead of a longitudinal approach, with a higher risk of drop-out, interviews might work better to get in-depth insights in the motivations of these low-motivated students.

Future research may investigate the role of classmates in students’ motivation, as this is still unclear. At the group level, relatedness with classmates has the smallest, but still significant relationship with intrinsic motivation. This indicates that having a good relationship with classmates is important for intrinsic motivation, but less important than the satisfaction of the other needs. However, from a within-person perspective, we find that the relationship between relatedness with classmates is even negative for a considerable amount of students. This difference cannot be explained by having SEN. It may be that it matters who the classmate(s) is/are who a student feels related to and how motivated this classmate is: students can positively influence each other (Gremmen et al., 2017), but it might also be that students demotivate or distract each other. Another option might be that not feeling related to any classmate does not matter for a student’s intrinsic motivation. Hence, it would be interesting to research friendship networks and their influence on motivation.

To conclude, this research reveals that it is important to conduct research on intrinsic motivation and basic psychological need satisfaction from the within-person perspective. Motivation and need satisfaction mainly fluctuate between students and lessons. Besides, individual differences in the relationships between need satisfaction and intrinsic motivation are found. This marks the importance of taking a within-person perspective, and studying both concepts at several lessons and individually. Students’ SEN play a role in those individual differences, but do not explain all differences between individual students. Future research may investigate other student characteristics to explore where individual differences come from. In-depth case studies may reveal even more information about how these processes work for individual students.

Implications for practice

At the group level, all needs have a significant positive effect on intrinsic motivation. Competence and relatedness with the teacher seem to be the most important, which stresses the importance of didactical skills and the quality of the teacher-student relationship. Teachers can support competence by providing structure, for example giving appropriate feedback and challenging activities. Teachers can support relatedness by showing involvement, for example showing warmth and respect (Niemiec & Ryan, 2009). However, individual differences in the importance of the needs for intrinsic motivation exist, so a teacher must take an individual perspective towards students. ESM may be a way to do this, as it can help teachers to address these individual differences in the importance of the needs, and help a student and his/her teacher to find out what is important for the motivation of that specific student.

Satisfaction of all basic psychological needs are positively related to intrinsic motivation of students with SEN, but competence and relatedness with their teachers seem to be more important. Therefore, if teachers want to pay extra attention to the motivation of these students, they should focus on providing structure and showing involvement. A good teacher-student relationship as well as didactical skills such as structuring a lesson, step-by-step instructions and adaptive feedback are then important teaching characteristics. However, teachers in the Netherlands are less able to use their didactical skills towards students with SEN (Smeets et al., 2019). It is important for teacher education and professionalization to focus on these skills.