Is problem-based learning associated with students’ motivation? A quantitative and qualitative study

In this study, a mixed-method design was employed to investigate the association between a student-centred, problem-based learning (PBL) method and law students’ motivation. Self-determination theory (SDT) states that autonomous motivation, which is associated with higher academic performance, can be reached when there is fulfillment of three psychological needs: autonomy, competence and relatedness. PBL aims to trigger autonomous motivation. In Study 1, 85 PBL law students (37% male; Mean age = 21.99 years) and 69 law students of a lecture-based, non-PBL program (39% male; Mean age = 22.72 years) filled out the Self-Regulation Questionnaire and an adapted version of the Work-related Basic Need Satisfaction Scale in order to measure autonomous and controlled motivation and perceived autonomy, competence and relatedness. In order to compare both groups, two MANOVAs were conducted and results showed differences neither in autonomous and controlled motivation, nor in feelings of autonomy and competence. However, PBL students experienced more relatedness. Additionally, in Study 2, focus-group discussions that were conducted indicated that PBL contains both autonomy-supportive and controlling elements, which might explain why no differences were found in perceptions of autonomy and autonomous and controlled motivation between PBL and non-PBL students. Furthermore, students reported that tutorial groups in PBL contributed to feelings of relatedness.


Introduction
Low graduation rates and high student dropout are two major issues that universities in higher education face all over the world. On average, 30% of the students enrolled in tertiary education leave without a degree (Organisation for Economic Co-operation and Development 2013). In the Netherlands, these are serious issues, especially in law schools compared with other disciplines (e.g. the graduation rate in Bachelor's degrees in 3 years is 25% in Dutch law schools compared with 32% of students in general; Central Bureau for Statistics 2014). A construct that is often associated with better academic achievement and graduation rates is students' motivation. For example, students' motivation highly correlates with academic achievements, such as grade point average (GPA; Richardson et al. 2012) and less intrinsically-motivated students are more likely to drop out (Vallerand et al. 1997). Hence, increasing and maintaining students' motivation in higher education are desirable. The design of a learning environment could help in this regard. Problem-based learning (PBL), a student-centred instructional method, aims to stimulate motivation. More specifically, one of the objectives of PBL is to foster intrinsic motivation in students (Barrows 1986;Hmelo-Silver 2004;Norman and Schmidt 1992).
In the present research, we explored whether PBL can positively affect students' motivation by conducting a quantitative study (i.e. Study 1: a comparison between a PBL and a non-PBL student cohort) as well as a qualitative study (i.e. Study 2: focus group discussions). These studies were conducted within a Dutch law school, because study progress issues are worrisome especially among Dutch law students. Self-determination theory (SDT), a wellknown theory of motivation by Deci and Ryan (2000), was used as the theoretical framework. SDT has been applied to the learning context, and components of SDT are much in line with the instructional method PBL (cf. Black and Deci 2000), as discussed later.

Theoretical background
Self-determination theory According to SDT, three basic psychological needs, namely, autonomy, competence and relatedness, are to be satisfied in every individual in order to stimulate psychological growth and well-being. Autonomy refers to having internal control over study activities and the learning process. Competence refers to the feeling of being capable of successfully performing study-related activities. Finally, relatedness refers to the need to feel warmth and support of others, such as teachers and fellow students Ryan and Deci 2000). As mentioned, SDT has been applied in the learning context to mean that, when the learning environment satisfies the three basic needs, students are more likely to become intrinsically motivated to learn (Katz et al. 2009).
Satisfaction of basic psychological needs determines the level of self-determination that is experienced. In SDT, a self-determination continuum is proposed consisting of different types of extrinsic motivation that move beyond the classic distinction between intrinsic versus extrinsic motivation. In the classic distinction, extrinsic motivation is often seen as detrimental for learning performances. However, not all types of extrinsic motivation hamper learning performances, depending on the amount of autonomy that is experienced . Instead, the distinction between different types of motivation can better be expressed by the differentiation between autonomous and controlled motivation. In autonomous motivation, self-determination is high. Autonomously motivated individuals act upon the activity because it is fun or interesting (i.e. intrinsic motivation) or because it enables personal development (i.e. identified motivation; Deci and Ryan 2000;Ryan and Deci 2000). Although the latter reason is extrinsic (i.e. the activity is not undertaken because it is interesting in itself), it is completely accepted and integrated with the self. In contrast, controlled motivation represents the kind of motivation in which selfdetermination is low. Students study because they experience pressure, such as trying to obtain a reward or avoiding punishment (i.e. external regulation) or to avoid feelings of shame and experience feelings of pride (i.e. introjected regulation).
Previous studies indicated positive relations between autonomous motivation and learning behaviour. A meta-analysis by Taylor et al. (2014) demonstrated a moderately strong, positive relation between autonomous motivation and school achievement. In that meta-analysis, studies from elementary school, high school and college were included. Furthermore, positive effects of autonomous motivation have been demonstrated for deeper learning and persistence in high school and college students of different educational programs (Vansteenkiste et al. 2004), better concentration and time-management in Chinese university students (Vansteenkiste et al. 2005) and lower dropout intentions in American high school students (Hardre and Reeve 2003). Controlled motivation, on the contrary, has been found to be negatively related to concentration and time-management and positively related to undesirable study behaviour, such as performance anxiety and dropout (Vansteenkiste et al. 2005).

Where problem-based learning meets self-determination theory
Considering the positive effects of autonomous motivation on learning outcomes, this type of motivation is desirable in students. Therefore, PBL specifically aims to stimulate students' intrinsic or autonomous motivation (Barrows 1986;Hmelo-Silver 2004;Norman and Schmidt 1992). PBL consists of three phases: the initial discussion, the self-study phase and the reporting phase. In the initial discussion, a collaborative discussion of a realistic problem (e.g. description of a real-life situation) takes place at the start of the learning process. Based on common sense and prior knowledge, students try to explain the problem. With the problem as the starting point of the learning process, knowledge of the topic addressed is limited and students collaboratively formulate questions about to-belearned aspects of the problem, called learning issues. In the second PBL phase, self-study, students individually search for and study relevant literature sources in order to answer the learning issues. After self-study, students return to the tutorial group to discuss the studied literature and address the learning issues together (i.e. the reporting phase). A tutor is present during the initial discussion and reporting phase. A tutor guides the process, for example, by intervening when students focus too long on irrelevant issues. He or she asks in-depth questions to make sure students themselves elaborate on course material, instead of providing them with factual knowledge (Barrows 1996;Loyens et al. 2012;Schmidt 1983). One could argue that several aspects of PBL encourage feelings of autonomy, competence and relatedness and subsequently students' autonomous motivation.
Students' needs for autonomy can be stimulated when students are provided with choice and when they can take control of their own learning . It is assumed that PBL stimulates students' autonomy in different ways. Because of its student-centred nature, students take control of their own learning, whereas the tutors have a facilitating role. The facilitating or guiding role of teachers in student-centred learning is assumed to support students' need for autonomy in SDT (Black and Deci 2000). Furthermore, PBL offers choice to students because of its emphasis on self-regulated learning. For instance, students formulate learning issues by themselves instead of receiving fixed learning issues from the tutor. Further, students choose and select their own set of literature sources, which stimulates autonomy. An empirical study by Wijnia et al. (2015) showed that having students select literature resources resulted in higher autonomous motivation scores than when students receive mandatory literature sources from an instructor in a PBL setting. The amount of autonomy increases when students are progressing in the academic program in PBL. For example, first-year students receive more guidance (e.g. more tips in providing literature and active scaffolding by the tutor) than third-year students.
Competence is experienced when students feel successful in a study task. Providing positive, informational feedback is one way to contribute to this (Deci et al. 2001). In PBL, the tutor provides formative feedback on how students function in tutorial group meetings (i.e. preparation for and participation in the reporting phase). Another way to anticipate feelings of competence is by providing problems that are based on real-life situations that need to be explained or solved. These 'authentic', realistic tasks can help students to feel more competent and confident in handling situations that will encounter in real-life and later in their profession (Dunlap 2005). It is likely that, if students feel that they can handle those types of situations, this will make them feel more confident and hence contribute to the second need of SDT, competence.
Regarding the third need, feelings of relatedness have a positive impact on students' intrinsic motivation (Ryan et al. 1994;Sheldon and Filak 2008). Students want to feel connected to and feel warmth from significant others which, in the learning context, involves relationships with teachers and fellow students. In a small, collaborative group setting (10-12 students), it is easier for students to build friendships which is helpful to increase their feelings of relatedness. In line with this assumption, PBL students were found to perceive collaboration in the tutorial groups as motivating (Wijnia et al. 2011). Additionally, in PBL, a tutor is present during small-group meetings. Because the groups are small, the tutor is able to give more individual support when needed and show interest in all students, which can stimulate feelings of relatedness as well.

Problem-based learning and motivational outcomes
Several studies of the PBL's effect on student motivation have been conducted. In these studies, a comparison was made between PBL students and students in a more traditional (mostly lecture-based) setting. Some studies indicated that PBL students scored higher on several motivational aspects, such as intrinsic goal orientation and enjoyability (Sangestani and Khatiban 2013;Sungur and Tekkaya 2006), which are important aspects of autonomous motivation. Another study found positive effects of PBL on self-efficacy (Liu et al. 2006). As mentioned earlier, when students feel more confident and competent, they are more likely to experience intrinsic or autonomous motivation. However, other studies show no differences for motivational outcomes between PBL and non-PBL students (Galand et al. 2010;Loyens et al. 2009;Wijnia et al. 2011). For example, Wijnia et al. (2011), using the SDT framework, found no differences for autonomous and controlled motivation. Similarly, Galand et al. (2010) found no differences for mastery and performance goals, which are constructs that share close associations with autonomous and controlled motivation, respectively .
A difference between the studies that found positive effects of PBL for motivational aspects (Liu et al. 2006;Sangestani and Khatiban 2013;Sungur and Tekkaya 2006) and studies in which no differences between these student groups were found (Galand et al. 2010;Loyens et al. 2009;Wijnia et al. 2011) is the length of implementation. In studies in which PBL positively related to motivation, PBL was often implemented for a short period of time (e.g. 15 days, Liu et al. 2006;6 weeks, Sungur and Tekkaya 2006;one semester, Sangestani and Khatiban 2013), while a curriculum-wide implementation of PBL was investigated in the studies in which no differences were found (e.g. Galand et al. 2010;Loyens et al. 2009;Wijnia et al. 2011). Why no differences in motivation were found in studies conducted with existing PBL curricula is puzzling. The need satisfaction of SDT (i.e. autonomy, competence and relatedness) was not taken into account in the PBL effect studies for motivation outlined above. Yet, the three needs are connected to several aspects of PBL (e.g. feelings of autonomy in PBL because of student-selection of literature), making the SDT an interesting framework for studies of PBL and motivation. In order to learn more about students' motivation in curriculum-wide PBL implementations, more insight into the relation between PBL and the need satisfaction is needed. The present study investigated students' motivation in a Dutch Law School, where a curriculum-wide implementation of PBL has taken place, and specifically focused on the role of PBL characteristics in students' perceptions of the three psychological needs.

The present study
This research consisted of two studies: a quantitative and a qualitative study. Two research questions were addressed in Study 1: 'What are the differences between PBL and non-PBL students regarding perceived autonomy, competence, and relatedness?' and 'What are the differences between PBL and non-PBL students regarding autonomous and controlled motivation?' In order to answer these questions, we conducted a quasi-experimental study in which third-year PBL law students and third-year law students of a lecture-based (i.e. non-PBL) method were compared on their self-reported autonomous and controlled motivation, as well as their perceptions of need satisfaction in their learning environment. Regarding the first research question, it was hypothesised that PBL students would perceive more feelings of autonomy, competence and relatedness. PBL is assumed to foster these three needs, because of certain characteristics that are present in this environment, such as students' selection of literature (i.e. for autonomy), use of real-life problems (i.e. for competence) and collaborative working in small groups (i.e. for relatedness). In turn, satisfaction of these needs in PBL is assumed to foster autonomous motivation and diminish controlling motivation. Therefore, with regards to the second research question, it was hypothesised that autonomous motivation would be higher among PBL students and that controlled motivation would be lower compared to non-PBL students.
In order to elaborate findings regarding the three needs, motivation and PBL, Study 2 followed up findings of the Study 1 by conducting focus-group discussions on the role of motivation and the three needs in PBL. Focus groups are discussion groups concerning specific questions and are helpful in exploring quantitative data (Kitzinger 1995).

Learning environment
The Dutch law program under study consists of a 3-year Bachelor program. In September 2012, all first-year law students who enrolled at the Dutch law School at the university under study started in the PBL program. Students who had already enrolled in the Dutch law school before September 2012 followed the Bachelor's program in a more traditional, lecture-based instructional environment. The differences between the two educational programs are described in Table 1.
The Dutch law study program in the lecture-based program consisted of four eight-week periods with two to three parallel courses. Lectures were emphasised as the main instructional method and, hence, students could attend multiple lectures each week in which a teacher transmitted information. Some courses offered a weekly work group in which a teacher explained and discussed a particular law case regarding the topic of the given course. Both the lectures and the majority of the work groups were non-mandatory. Examination weeks were held four times each academic year at the end of each 8-week period. During these examination weeks, multiple courses were examined.
In September 2012, the Dutch law program shifted from traditional, lecture-based learning to PBL. Teachers were trained to adapt their teaching style from a teacher-centred, directive style to a more guiding, facilitating role. Additionally, new tutors were hired and trained. Further, training for changing courses and creating problems was provided. The new PBL program is different from the traditional program in several ways. The PBL program consists of eight sequential courses each academic year, which means that courses are not offered in parallel anymore. Each course takes 5 weeks (i.e. block) and ends with a written examination instead of four examination weeks within the academic year. The tutorial group meetings, which are held twice a week, are considered an important element in the PBL program. The groups consist of 10-12 students and a tutor. The group composition changes each block. Each five-week course consists of eight problems, all addressing different, but related, topics within the course. To give an example, one of the problems in a criminal law course could focus on self-defense. A (fictitious) news article could serve as the problem, describing a realistic situation in which a man is using selfdefense when he is attacked. During a tutorial meeting, the reporting phase of a problem and the initial discussion of a new, subsequent problem take place. In the example of the problem regarding self-defense, students discuss in the initial phase whether they think that the man had the right to defend himself, ending with questions (i.e. learning issues) when self-defense applies. Between these meetings, students have 2-3 days of self-study in which they prepare themselves for the upcoming meeting. They search for and select information from different sources, such as text books, laws and jurisprudence to use in addressing the learning issues. In the reporting phase, students collaboratively discuss the studied materials and learning issues. Law students in general need to learn how to reason about legal cases. The problems used in PBL help students to think about realistic situations in which they need to apply what they have learned. In the Dutch law system, rules and principles are applied more often than comparisons with prior case law, as is the case in UK-and US-law. Besides the tutorial meetings and self-study, students participate in practical courses that help them to learn how to apply the learned knowledge. For example, students learn to plea in front of a judge and a lawyer with a realistic law case. Students earn study credits when passing the assignments of these courses. Further, non-mandatory lectures are provided by teachers two or three times a week, to expand the knowledge that is acquired during the tutorial meetings.

Participants
In the current study, participants were third-year Dutch law students of two cohorts. A comparison between both student cohorts took place, and hence participants were students from the first cohort of the PBL program and students from the last cohort of the traditional, lecture-based program (i.e. non-PBL students). Eighty-five PBL students (37% male) and 69 students of the lecture-based, non-PBL program (39% male) participated. Mean age was, respectively 21.99 years (SD = 2.02) and 22.72 years (SD = 3.15). Students in both cohorts differed neither with respect to age, t(152) = 1.76, p = 0.081, nor gender, x 2 (1) = 115, p = 0.735. The male/female ratio in both groups is representative for Dutch Law Schools (Central Bureau for Statistics 2014).

Materials
Students' perceptions of autonomy, competence and relatedness and their autonomous and controlled motivation were measured with two existing questionnaires. It was explicitly stated that students should base their answers on their experiences of the entire Bachelor's program (i.e. the first 3 years of the academic program) rather than solely on experiences of the course in which they participated at the time when they received the questionnaire.

Satisfaction of needs
The way in which students perceive autonomy, competence and relatedness in their learning environment was measured with the Work-related Basic Need Satisfaction Scale Therefore, some adjustments were made in order to fit the items of the questionnaire to a learning environment (e.g. the word 'work' was replaced by 'study'). The adapted version of the W-BNS contains 18 items that are rated on a five-point Likert scale (1 totally disagree to 5 totally agree). The questionnaire consists of three scales with six items each. Table 2 presents questionnaire characteristics of the adapted WBN-S.

Autonomous/controlled motivation
The Self-Regulation Questionnaire (SRQ; Vansteenkiste et al. 2009) was used to measure autonomous and controlled motivation. In this questionnaire, students are asked to rate different reasons for studying on a five-point scale ranging from 1 (not important at all) to 5 (really important). The SRQ contains a total of 16 items divided into four scales: external regulation, introjected motivation, identified motivation and intrinsic motivation. Based on previous research (e.g. Vansteenkiste et al. 2009), the four scales were combined into two types of motivation, namely, controlled motivation (i.e. average scores of the scales introjected motivation and external regulation; Cronbach a = 0.85) and autonomous motivation (i.e. average scores of the subscales identified motivation and intrinsic motivation; Cronbach a = 0.89). See Table 3 regarding questionnaire characteristics.

Procedure
A cohort comparison between PBL and non-PBL students was carried out within one university. The PBL students entered the law school in their first year in September 2012 and the non-PBL students entered their first year in September 2011. Both student groups participated when they were in their third year. Hence, students of the PBL cohort filled out the questionnaires in April 2015 and the non-PBL students a year earlier, in April 2014. In this way, students of the non-PBL and the PBL programs could be compared while they were in the same phase of the academic program (i.e. third year).
Because of the shift of the educational program, there were some changes in the course order as well. Non-PBL students received the questionnaires during a non-mandatory lecture of the course Business and Corporate Law. One of the authors handed out the questionnaires to the students and collected them after they had been completed. In the PBL cohort, questionnaires were distributed by tutors during the final (mandatory) tutorial meeting of the course Philosophy of Law. Completing the questionnaires took students about 10-15 min. Afterwards, tutors collected the questionnaires and handed them over to one of the authors.

Statistical analysis
To investigate the effects of PBL on students' perceptions of the satisfaction of the three psychological needs and their motivation, two separate Multivariate Analyses of Variance (MANOVAs) were conducted. The first MANOVA focused on the three needs. Instructional method (i.e. PBL vs. non-PBL) served as the between-subject factor and satisfaction scores for the three needs in the learning environment (i.e. autonomy, competence and relatedness) were dependent variables. The second MANOVA concerned scores on the SRQ. Again, instructional method (i.e. PBL vs. lecture-based) served as the betweensubjects factor and motivation scores (i.e. autonomous and controlled motivation) as dependent variables. Effect sizes were expressed in partial eta squares (i.e. partial g 2 ), and were indicated as small, medium, or large effects when values were .01, .06, and .14 respectively (Richardson 2011).

Results
Mean scores for both student cohorts on the adapted version of the W-BNS and the SRQ are given in Table 4. First inspection of scores for the three needs showed that they were all rather high, especially scores for competence. Scores on autonomous motivation were higher compared with controlled motivation in both student groups. Table 5 provides correlations between all variables. The psychological needs were positively and highly correlated with autonomous motivation, with exception of relatedness (i.e. nonsignificant correlation). Correspondingly, controlled motivation negatively correlated with perceived autonomy and competence. Again, no correlation with relatedness was present. Before conducting the MANOVAs, assumptions were checked and met (e.g. normality of residuals of dependent variables, Box's test for homogeneity of covariance matrices was nonsignificant for the first and second MANOVA, respectively p = 0.175 and p = 0.109). The first MANOVA for the three basic needs autonomy, competence and relatedness showed a medium effect for instructional method, Pillai's Trace (V) = 0.06, F(3, 150) = 3.31, p = 0.022, partial g 2 = 0.06. To follow up this MANOVA, separate ANOVAs were conducted. In order to reduce the chance of Type I error, a Bonferroni correction was applied and results were only considered significant when an alpha level of 0.017 was reached (0.05/3). Differences emerged between student groups neither for perceived autonomy, F(1, 152) = 1.60, p = 0.207, partial g 2 = 0.01, nor perceived competence, F(1, 152) = 0.04, p = 0.844, partial g 2 \ 0.01. However, a significant difference emerged for the satisfaction of the need for relatedness, F(1, 152) = 6.88, p = 0.010, partial g 2 = 0.04 (i.e. small effect) in favour of the PBL students. The second MANOVA for autonomous and controlled motivation showed no effect of instructional method on students' motivation, Pillai's trace (V) = 0.01, F(2, 151) = 0.36, p = 0.696, partial g 2 = 0.01.

Discussion
PBL and non-PBL students did not differ in their feelings of autonomy and competence in the learning environment. These results were unexpected because it was believed that PBL would stimulate autonomy (e.g. choice in literature sources) and competence (e.g. work on realistic tasks) more than the traditional program. Further, it was found that feelings of relatedness were higher in PBL students than in non-PBL students, meaning that PBL students experience more support by others such as teachers and peers. There was no correlation, however, between autonomous motivation and relatedness, and between controlled motivation and relatedness. Despite higher scores on relatedness, students' Scores on both questionnaires range from 1 to 5  . Possible explanations are discussed in the general discussion below. Results further demonstrated no differences between PBL and non-PBL students in their autonomous and controlled motivation. These findings were not in line with findings of Sangestani and Khatiban (2013) and Sungur and Tekkaya (2006), which demonstrated positive effects of PBL on student motivation, but they were in line with results reported by Galand et al. (2010), Loyens et al. (2009) and Wijnia et al. (2011). While the studies that found positive outcomes implemented only a short-term PBL intervention, the other studies (Galand et al. 2010;Loyens et al. 2009;Wijnia et al. 2011), as well as the current study, were conducted in existing PBL curricula. Introducing students to a short PBL intervention might only influence their motivation, because the method is completely new to them. Conducting the studies with existing curricula is more ecologically valid. Furthermore, correlations indicated that perceived autonomy and competence were positively and moderately to highly correlated with autonomous motivation and negatively and moderately correlated to controlled motivation (see Table 5). Because scores on competence and autonomy feelings were high in both PBL and non-PBL students, the absence of significant differences between groups on autonomous and controlled motivation become clearer.
Considering that most of the findings were not in line with the hypotheses, with the exception of higher relatedness scores among PBL students, a follow-up study with focusgroup discussions was conducted to add to and explain these findings. The focus-group discussions attempted to elaborate elements in PBL that can satisfy or thwart the three needs and the motivating and demotivating elements in PBL. Specifically, students discussed which PBL characteristics influence their feelings of autonomy, competence and relatedness in order to acquire more understanding of the lack of differences regarding autonomous and controlled motivation and of perceived autonomy and competence.

Study 2: focus-group discussions
As we were interested in the relation between different aspects of PBL and the components of SDT, two focus-group discussions with PBL students took place. During focus groups, students give their opinions on certain topics and collaboratively discuss them. Findings from focus group discussions add to data from quantitative studies (Kitzinger 1995) and offer more understanding of why certain results showed up. During the focus groups, students elaborated PBL characteristics and whether these were experienced as motivating or demotivating, as well as the degrees of autonomy, competence and relatedness that they experienced in PBL and which elements in PBL contributed to this.

Participants
Third-year Dutch PBL law students were recruited and informed about the process of the focus groups and that the discussion would focus on PBL. They were guaranteed that their contribution would be reported anonymously. In total, 13 students volunteered to participate and they were assigned to one of two focus groups, depending on the time of their tutorial meeting, because the focus group took place prior to or after their meeting. PBL Learning Environ Res (2018) 21:173-193 183 students who participated in the focus groups were also involved in the quantitative study and filled out the questionnaires on autonomous and controlled motivation. The first group consisted of five students (one male, four females) and the second group consisted of eight students (three males, five females). The focus groups were held on one day, directly before or after one of the tutorial meetings in the final course of the third academic year (June 2015). Students were recruited from different tutorial groups.

Procedure
The first author acted as interviewer in both groups. She asked the questions, took notes and made sure certain topics were covered in the discussion. The first open-ended question was: ''Which aspects of PBL do you consider motivating and which aspects do you consider demotivating?'' Additionally, the interviewer introduced the three psychological basic needs of SDT briefly. Then the following three questions were asked: ''Do you have the feeling there is autonomy in PBL and which characteristics of PBL contribute to this feeling?'', ''Do you feel competent in PBL and which characteristics of PBL contribute to this feeling?'' and ''Do you experience relatedness in PBL and which characteristics of PBL contribute to this feeling?'' Students were instructed to answer freely and discuss each other's opinions. The authors agreed beforehand on the need to address certain topics concerning the most important characteristics of PBL, such as the tutor, the problems used in PBL, collaboration, self-regulated learning and connection with practice. Furthermore, topics concerning the implementation of PBL in the curriculum under study, such as the lectures, needed to be addressed. When these topics were not addressed spontaneously, the interviewer asked students' opinion about the role of the particular topic with respect to their motivation/demotivation. Both focus-group discussions took about 60 min and were recorded.

Analysis
The first focus-group discussion was transcribed literally. Because of a technical problem, recording the second discussion failed. Therefore, the interviewer directly wrote the discussion down after it took place, based on the written notes and memory. This summary of the discussion was analysed instead. Statements in the transcriptions were classified under one of five categories, which are based on SDT: motivating aspects, demotivating aspects, autonomy, competence, and relatedness. One of the authors and an independent rater both categorised all statements. There was substantial agreement between raters (j = .80) and discrepancies were resolved through discussion.

Motivating aspects
Overall students experienced PBL as satisfying. The structure that PBL offers, such as a period of self study prior to a group discussion, and the fact that courses are offered in succession, were pleasant. The tutor and the problems used, which are specific characteristics of PBL, were perceived as motivating, as long as they met certain conditions. Students were enthusiastic about the tutor when he/she showed interest, had expertise and was actively involved in the group (i.e. asking in-depth questions and helping when students discuss irrelevant information). In general, students were positive about the problems used in PBL. For example, students indicated that, when the problem is used to apply the acquired knowledge in the reporting phase, this is enjoyable: I think it is motivating in PBL that the case [the problem] triggers you to find things out. FG1, S2 […]. That is motivating to me, when at the end of the reporting phase you understand how it [the problem] in a realistic situation works. FG1, S4 It is motivating when I get the feeling the tutor understands the learning material […]. FG1, S4

Demotivating aspects
There were also some perceived demotivating aspects of PBL. For example, in students' opinions, the initial discussion was sometimes redundant and could be shortened (e.g. formulating the learning issues more directly without a discussion). Moreover, if the initial phase of PBL lacks discussion, students were demotivated. When the topic of the problem is too abstract or too far removed from the students, they lack prior knowledge and experience difficulties discussing the topic: For example, in the course Philosophy of the Law, one can take different perspectives, which makes discussion possible. But for example in the course (Dutch) Civil Procedural Law, all we need to know is written down in the Civil Code, so you don't really have an opinion about it. This makes it hard to enter discussion in the initial discussion. FG1, S1 Some specific elements of PBL that were earlier described as motivating (i.e. tutor and problems) can also be considered demotivating under other conditions. For example, a tutor was considered very demotivating when he/she was passive during the meetings (i.e. hardly asking questions and being inattentive in the discussion). Further, problems that were too long or similar to previous problems were also unsatisfying: It is really demotivating when a tutor is passive and does not intervene in the discussion when necessary and gives us the feeling he/she doesn't understand what is discussed in the group. FG1. S4.
Another aspect of the educational system that caused a lot of discussion in the focus groups was the mandatory attendance requirement for tutorial meetings. In the PBL curriculum under study, students are required to be present during the tutorial meetings. They are allowed to miss only one meeting per course and this needs to be compensated with a compensatory assignment. Although understanding the importance of attendance in the tutorial meetings, students felt that this rule is too strict. Lectures were also perceived as being demotivating, especially when they are not interactive. Students argued that there were too many lectures in a row, making it difficult to stay focused (approximately 4-6 h): Lectures are good when the lecturer let's students participate, but only a few lecturers do this […]. FG1, S2. Learning Environ Res (2018) 21:173-193 185 Autonomy When students were asked directly whether they experienced autonomy in PBL, the majority reported feelings of low autonomy. Factors that contributed to this were the mandatory attendance presence, lack of choice in courses and not being able to select their own tutorial group, because students are randomly assigned to their tutorial group. However, students did also mention some autonomy-supportive elements in PBL, such as choice in literature sources and room for their own discussions in the tutorial meetings, without interruptions of the tutor. Interestingly, students were unsatisfied with these autonomy-supportive aspects of PBL: I think it is demotivating that teachers want you to read multiple literature sources during one course. They recommend five to six books, but you will never study all of them.
[…] I think this is confusing. FG1. S2 It would be nice if the tutor guides more often in a way that he or she would make it more clear what we need to know during the discussion. FG1.S4 Further, the required preparation for every meeting, which is more a controlling element in PBL, served as an incentive to study. Students study on a regular basis that way.

Competence
In general, students felt competent during their study. Both nonspecific PBL elements (e.g. achievements in form of grades) as specific PBL elements (e.g. the phases of PBL) contributed to feelings of competence. During the phases of PBL (i.e. initial discussion, selfstudy, and reporting phase), students first activate their prior knowledge, then individually study the material, and afterwards discuss the material collaboratively. It seems that being actively involved in the learning process contributes to feelings of competence: I believe that PBL offers the possibility to really understand the material, because you can ask a lot of questions and you can discuss [about the material]. So you'll know whether you get it or not and this gives a feeling of certainty before you enter your examination. Because you know you have discussed all of it. FG1.S4 As mentioned before, students like to apply the learned knowledge to the problem. In addition to the fact that this is motivating, connecting theory and practice helps to create feelings of competence and helps students to build coherent understanding of the material: […]. You can apply the theory you learned on a practical case [when working with the problem]. Otherwise it [learned course material] stays so abstract. FG1.S4

Relatedness
All students indicated that they felt connected with others. The most important PBL factor that contributes to this is the tutorial group, because students get to know each other in the meetings. Additionally, students feel that the tutor is approachable in PBL, and hence they are more likely to ask questions or start a conversation with him/her: You know a large number of law students by now, because there are different students in your tutorial group every course. I really like that, meeting so many new people. FG1. S5

Discussion
Results of the focus groups analysis showed that PBL students indicated presence of both motivating and demotivating elements in the learning environment. In general, students were satisfied with PBL. Especially the process of PBL (i.e. self-study before discussion of the material), sequential courses (i.e. one course for 5 weeks, ending with an examination) and an active tutor were motivating. Yet, there also were some perceived demotivating aspects in PBL, such as the initial discussion, a passive tutor and mandatory attendance.
Other statements in the focus-group discussions concentrated on the three psychological needs according to SDT . Students experienced some autonomy, but also felt that they were controlled by certain PBL elements such as the mandatory presence and required preparation. Feelings of competence were attained by specific PBL elements (i.e. realistic problems) and non-specific PBL elements (i.e. grades). Further, the tutorial meetings with fellow students contributed to relatedness.

General discussion
As motivation is important for academic success and study progress (Richardson et al. 2012;Vallerand et al. 1997), motivation needs to be stimulated in students. PBL is an instructional method that aims to foster intrinsic motivation (Barrows 1986;Hmelo-Silver 2004;Norman and Schmidt 1992). Hence, the present study investigated the relation between PBL and Dutch law students' motivation using a mixed-methods design. SDT was used as a theoretical framework to investigate the claim that PBL can indeed foster students' intrinsic or, in SDT-terms, autonomous motivation. Study 1 involved a comparison between students of a PBL cohort with students of a lecture-based cohort (i.e. non-PBL) in terms of their perceived feelings of autonomy, competence and relatedness in the learning environment and their autonomous and controlled motivation. Perceptions of students' need satisfaction were included because these needs are important for the experience of motivation (see Deci and Ryan 2000). Results showed no differences in feelings of autonomy and competence, but PBL students experienced more relatedness in their learning environment. Further, no differences were found for both types of motivation. In Study 2, qualitative data concerning the role of PBL for motivation and need satisfaction (i.e. autonomy, competence and relatedness) were collected with focus-group discussions to follow up the results of Study 1.

Autonomy, competence and relatedness
SDT states that, when the social context of a learning environment satisfies the needs for autonomy, competence and relatedness, students become autonomously motivated . Previous studies investigating differences between PBL and non-PBL students' motivation did not include students' perceptions of this need satisfaction. Examining need satisfaction might be insightful because these needs are important antecedents of motivation . It was expected that feelings of autonomy, competence and relatedness would be stimulated more in PBL than in a traditional, lecturebased curriculum. Yet, results were not completely in line with these expectations.
With regard to autonomy, no differences were found between PBL and non-PBL students. In the focus-group discussions, it appeared that there were a number of autonomy- Learning Environ Res (2018) 21:173-193 187 supportive elements present in PBL (e.g. some choice in literature), but also there were controlling elements (e.g. lack of choice in tutorial group composition). One can assume that, in the non-PBL environment also, both autonomy-supportive (e.g. choice in fellow students for collaborative assignments) and controlling elements (e.g. prescribed literature) were present. The presence of controlling elements in PBL and probable autonomy-supportive elements in a non-PBL environment could help to explain why no differences emerged for perceived autonomy. When asked directly during the focus-group discussions, students indicated low degrees of autonomy and high feelings of control. The main contributing factor to this feeling was mandatory attendance at tutorial meetings. However, one could argue that mandatory attendance does not refer to an autonomy-supportive or controlling element, but more to a structural element in PBL. Providing structure holds that students are offered clear instructions of what is expected of them (Jang et al. 2010), such as instructions about presence. In general, providing structure is beneficial for educational results relative to no structure in class (Jang et al. 2010). Yet structure can be offered in an autonomoussupportive way (i.e. discussing rationale, taking students' feelings into account), which is beneficial for students, or in a controlled way (i.e. no discussion of rationale, not taking students' feelings into account), which has a detrimental effect on students (Jang et al. 2010). It is possible that communication about mandatory attendance in the curriculum under study was perceived as controlling rather than autonomy supportive.
Moreover, although elements such as choice in literature sources and limited interferences of the tutor were intended to be autonomy supportive in nature, students were unsatisfied with these elements. It is possible that the amount of autonomy expected from students, with respect to literature selection for example, was too high, making students feel lost in the course material (Sierens et al. 2006). Kirschner et al. (2006) described this in terms of minimal guidance which, according to them, is harmful for learning. In PBL, the amount of instructions should be adapted to the level of the student (i.e. scaffolding; Schmidt et al. 2007). For example, novice students (e.g. first-year students) are provided more help in literature searches (e.g. more tips) compared to experienced students (e.g. third-year students), because novice students lack experience (Schmidt et al. 2007). Possibly, in the curriculum under study, students (even in their third year) experienced difficulties with respect to their responsibility for literature choices, resulting in feelings of uncertainty.
Considering the need for competence, students indicated that the phases of PBL help them in experiencing feelings of competence. PBL offers opportunities to rehearse course material, which make students feel confident about the learned material. Moreover, the discussion during the reporting phase helps students to create a rich understanding of the course material. Students indicated that the use of realistic problems also contributed to feelings of competence, which is in line with the study by Dunlap (2005). Real-life problems support a connection between theory and practice, leading to a better understanding about the material. Yet, non-PBL students also reported feelings of high competence in the learning environment. A first explanation is that some courses in the non-PBL curriculum also offered work groups in which students worked on a realistic law case, contributing to feelings of competence in non-PBL students as well. Second, non-specific PBL factors that contribute to feelings of competence, such as obtaining good grades, are common in both instruction types, explaining why no difference in competence showed up. Finally, students of both cohorts were third-year students and probably all experienced feelings of competence, because they all succeeded so far in their academic careers.
The only difference between PBL and non-PBL students was in feelings of relatedness. Specifically, PBL students reported feelings of higher relatedness compared with non-PBL students. Analysis of focus-group discussions demonstrated that this feeling can be explained by the opportunity to form peer connections in tutorial meetings. In PBL, students meet twice a week in a small (i.e. 10-12 students) tutorial group and the groups change each course. In PBL, students therefore get to know a large number of fellow students in this way. Alternatively, it is likely that large-scale, lecture-based curricula (i.e. traditional) create a sense of anonymity among students and are more impersonal. The teacher is less involved and more distant than in PBL.
Correlations between relatedness and autonomous and controlled motivation were nonsignificant. This finding was not in line with results of previous studies (e.g. Sheldon and Filak 2008) in which positive relations between feelings of relatedness and intrinsic motivation were demonstrated. Still, even though there is no relation with motivation, feelings of high relatedness are beneficial for other student outcomes such as student dropout. Tinto's (1975) model stresses the interaction between students and the academic environment and its influence on student dropout. If students are socially integrated in the academic environment, commitment increases, making it less likely that students voluntarily drop out of college (Tinto 1975). Social integration is the result of connections with peers and interactions with staff. Results of our study suggest that social integration is present in PBL more than in a non-PBL environment. Students feel related through smallscale tutorial groups in PBL, because they get to know one another in both a formal (i.e. collaborate on study activities) and informal (i.e. friendship) way. In addition, interaction with tutors in the groups contributes to social integration. This result is in line with findings of a study by Meeuwisse et al. (2010) which indicate that an active learning environment (i.e. such as PBL) fosters interactions with both teachers and students.

Autonomous and controlled motivation
It was anticipated that PBL students would report higher scores for autonomous motivation. However, Study 1 revealed no differences for autonomous and controlled motivation between the two student cohorts. But PBL and non-PBL students reported rather high autonomous motivation scores (M = 3.82 and M = 3.85, respectively, range 1-5). These results indicate that the claim that PBL can stimulate students' intrinsic motivation was not supported by our results. A first explanation has to do with the findings for the three psychological needs. No differences between PBL students and their non-PBL counterparts were found for perceived autonomy and competence. Correlations reflected a positive relation between perceived autonomy and competence with autonomous motivation, and a negative relation between perceived autonomy and competence with controlled motivation. Because scores on perceived feelings of autonomy and competence did not differ, it is not surprising that no differences were found for autonomous and controlled motivation.
Another possible explanation for why there were no differences between PBL and non-PBL students for autonomous motivation is that participation in our studies by third-year Bachelor's students took place at the end of the academic year. Apparently, all participants were enthusiastic about their study and were motivated to finish the Bachelor's program. In general, students who are autonomously motivated continue the academic program, while controlled motivated (or demotivated) students drop out at an earlier stage (e.g. Vansteenkiste et al. 2005;Vallerand et al. 1997). Nevertheless, third-year law students were chosen because these they had more experience with the academic program and curriculum (relative to first-year students), making their opinions rather valuable for the focus-group Learning Environ Res (2018) 21:173-193 189 discussions. Nevertheless, we anticipated that similar effects would have been found if first-year students were questioned. Results are in line with a study that was conducted with predominantly first-and second-year students of a PBL psychology program (Wijnia et al. 2011). In that study, similar to our results, no differences were found between PBL and lecture-based students for autonomous and controlled motivation. Therefore, we assume that the results can more likely be explained by the fact that no differences were found for the perceived needs of autonomy and competence.

Limitations, recommendations for future research and implications
The present study had some limitations. A first limitation is the participation of third-year students. It is likely that third-year students would be more motivated and confident about their study than first-year and second-year students, because they almost had finished the Bachelor's program. However, third-year students also were more experienced with the PBL program and therefore their opinions were valuable for the focus-group discussions. Second, non-PBL students filled out the questionnaire during a non-mandatory lecture, while the PBL students filled out the questionnaires during a mandatory meeting. It is likely that the students who were present during the lecture were highly motivated, which could have biased our results. Nevertheless, results are in line with previous studies conducted in existing PBL curricula (e.g. Galand et al. 2010). Further, administration of the questionnaires took place during different courses in both student groups because of changes in course order. Even though students were instructed to base their answers on the entire Bachelors' program, it cannot be ruled out that the content of the course had some sort of influence on the answers. Finally, with regards to the focus-group discussions, recording of one of the discussions failed. Even though the interviewer directly wrote down the content of the discussion, exact statements were missing for this group. Partly based on these limitations, we have some recommendations for further research. Although the main focus of the present study was the influence of PBL on student motivation, it would be interesting to conduct focus groups among non-PBL students as well. At this point, we can only make assumptions about which factors influence student motivation under traditional instruction. Further, the present study indicated that there was no correlation between perceived relatedness and either autonomous motivation or with controlled motivation. Further research is needed into why this relation is absent. Moreover, it might be valuable to connect dropout to motivation, especially feelings towards relatedness. Relatedness, which appeared higher among PBL students, might influence student dropout according to Tinto's model.
In this study, we used SDT as the theoretical framework. We realise that other motivational theories might be of interest as well, such as achievement goal theory or expectancy-value theory. However, in the current study, we were mainly interested in investigating whether PBL can indeed stimulate higher levels of intrinsic or autonomous motivation.
Both the quantitative and qualitative studies were conducted with Dutch law students, because potentially they could benefit most from improvements in motivation (with regard to low graduation rates and high dropout rates among Dutch law students) (Central Bureau for Statistics 2014). However, results are also insightful for other higher educational programs: student-centred instructional methods, based on constructivist learning theories, have received much attention over the past decades (Baeten et al. 2013) and these methods replace conventional lecture-based programs more and more in several disciplines (White et al. 2016). Because PBL can be considered an active and constructivist learning approach, findings of the present study for an activating learning approach and motivation are therefore important for other programs and disciplines as well.

Conclusion
The present study showed no differences between PBL and non-PBL students regarding autonomous and controlled motivation, as well as perceptions of autonomy and competence. Students in both educational forms were highly autonomously motivated and experienced feelings of autonomy and competence in their learning environment. This could be attributable to the presence of both autonomy-supportive and controlling elements in the PBL learning environment, although a difference in feelings of relatedness was found in favour of PBL. The small tutorial groups in PBL seemed to contribute to these feelings of high relatedness, as students get to know their peers and feel that their teachers are more approachable.