Promoting self-directed learning skills in a first-year problem-based university program with a self-development intervention

ABSTRACT

This study investigated whether an intervention aimed at self-development would stimulate the self-directed learning skills of university students within a problem-based learning program. In a randomized controlled trial study, 408 first-year psychology students were randomly assigned to an experimental or control condition. During their courses, all students received the problem-based program, including feedback from their tutors on professional behaviour and study skills. Students in the experimental group also received an extra self-development intervention including two sessions with a mentor, during which students were asked to reflect on personal learning goals. At the beginning and end of the year, students’ self-directed learning skills were measured with two questionnaires. Over the year, students in the experimental group showed a larger increase in their self-control, in their ability to monitor improvement during task performance, and in their ability to evaluate process and outcome after task execution. Also, their willingness to invest study effort had not decreased, in contrast to students in the control group. These findings show that a relatively simple intervention making students reflect more on their self-development can have a significant effect on a number of relevant self-directed learning skills, even when added to a problem-based learning university program that is already focused on stimulating self-directed learning.

KEYWORDS:

• Self-directed learning
• self-development
• problem-based learning
• intervention study
• higher education

Given the rapid technological and global changes in society, higher education is struggling to create educational situations for all the possible knowledge students need in their future. As a result, students need to develop skills that prepare them to be lifelong learners (Morris, 2019). Self-directed learning (SDL) is suggested as the main methodology of choice in this context (Fischer & Sugimoto, 2006). Knowles (1975, p. 18) describes SDL as ‘a process in which individuals take the initiative with or without the help of others in diagnosing their learning needs, formulating goals, identifying human and material resources, and evaluating learning outcomes’.

One educational format developed to foster SDL in students in higher education is Problem-Based Learning (PBL; Barrows, 1986). The curriculum in PBL is structured into thematic blocks with corresponding problem cases that trigger students in a study group to clarify and analyse the problem cases together and formulate learning objectives (Hmelo-Silver, 2004). Subsequently, students go home to independently study resources that are relevant for answering the learning goals, before returning to the group to discuss and fine-tune their acquired knowledge. Students learn the PBL-approach as a procedure at the start of students’ education, and this PBL-procedure gives the opportunity to develop deeper learning strategies that are needed to clarify and analyse problem cases, formulate learning goals, and do independent study (de Graaff & Kolmos, 2003; Wood, 2003).

In the context of PBL research, SDL is generally studied by looking at the deeper learning strategies that students obtain through the PBL-procedure, as well as the improved monitoring of their study progress (Loyens et al., 2008; Malan et al., 2014). When taking the definition of Knowles (1975) into account, this view on SDL in the PBL-context seems a bit limited, as it does not include ‘taking the initiative’, ‘diagnosing learning needs’ and ‘identifying human and material resources’. These aspects are especially important for self-development during lifelong learning. Self-development assumes that people are not only capable of applying appropriate learning strategies, but also of actively monitoring their own behaviours, seeking feedback, recognising effective behaviours to diagnose their learning needs, and specifying who and what they need (the resources) to reach their goals (London & Smither, 1999). The predefined context of PBL may not always facilitate learning all aspects of SDL necessary for self-development: the course contents and resources such as literature are provided by the educational institute and are seldomly based on choices from individual students. That raises the question to what extent a PBL-approach can also be successful in promoting the kind of SDL needed for self-development during lifelong learning.

Promoting self-directed learning

The definition of Knowles (1975) may suggest that SDL is a linear process that students go through. However, simply stimulating students to follow these steps will not always promote SDL. Learner attributes can interfere with or facilitate the process (Brockett & Hiemstra, 1991; Garrison, 1997). Attributes such as self-management skills, responsibility, prior knowledge, and motivation are shaped in the SDL process but also shape the SDL process. SDL is also situational: students’ self-directness may differ in different learning situations depending on learner attributes relevant for that situation (Candy, 1991). Therefore, in order to promote SDL, different opportunities and varying degrees of support by the instructor are required and should be balanced with students’ level of self-directedness.

Long (1989) provided an instructional model of SDL, in which the learners maintain active control of the learning process (psychological control) in balance with the degree in which learners get the freedom to determine learning goals, seek resources, set the mode of evaluation, and are given support (pedagogical control). Students with a high level of psychological control should receive low pedagogical control to stimulate SDL and vice versa. In this sense, more pedagogical control could facilitate first-year students that are generally lower in their ability of SDL because they have just started with their higher education.

There is also some evidence that shows that offering pedagogical control to first-year students within the PBL-procedure is beneficial for the development of their SDL, mostly in their use of deeper learning strategies (Loyens et al., 2008; Malan et al., 2014). For example, Malan et al. (2014) evaluated a new PBL curriculum and found that compared to a traditional teaching method, PBL facilitated the use of deep learning strategies. Most students also reported that these learning strategies provided more planning, motivation, insight and activity. However, in another study that looked into self-development, medical students participating in a PBL-program reported difficulties with personal reflection, goal generation, environmental strain, competing demands, and experienced problems with plan development and implementation (Li et al., 2010). Thus, setting and pursuing learning goals for a problem case does not necessarily seem to transfer to setting and pursuing goals for self-development in the context of lifelong learning.

So, the question is how to provide pedagogical control that promotes SDL required for self-development within a PBL-program. Reviews of intervention studies have shown that SDL (in the form of self-regulatory processes) can effectively be promoted in higher education (de Bruijn-Smolders et al., 2016; Eggers et al., 2021). However, most of these studies focused on strategies to improve self-regulation, and none focused on setting and pursuing goals for self-development. However, a recent intervention study, Colomer et al. (2021) reported about a self-development intervention to improve self-directed learning within a regular university program. They set up a longitudinal study over the course of four years, during which 51 Science students were stimulated in their self-directed learning by reflecting on the assessment of their self-directed competencies (twice a year), by receiving personal tutoring (three to four times a year), and by making a portfolio reflecting about their development (once a year). On a self-report measure, they did not find changes in factors like self-reflection on learning and self-regulation, but they did find that students’ ability to analyse their current situation seemed to improve, especially during the first two years. So at least for some dimensions of SDL, a self-development intervention where students work with personal goals, reflect and discuss their development with tutors seems promising. However, the study of Colomer et al. (2021) did not use a control group, so it is difficult to weigh the added value of their intervention. Moreover, the program was not based on PBL, so it remains unclear whether similar results may be obtained within a PBL environment.

The current study

Although there is evidence that a PBL environment facilitates some aspects of SDL, the question is whether it could also promote the aspects of SDL needed for self-development during lifelong learning. At the same time, there is some preliminary evidence that adding a self-development intervention to a regular university program may benefit at least some self-directed learning aspects related to self-development. Therefore, the aim of this study was to find out whether offering a self-development intervention may further stimulate the self-directed learning of first year students within a PBL-based university program. The intervention used a similar approach as Colomer et al. (2021), as it stimulated students to work with personal goals, reflect on their experiences and discuss their progress with a mentor.

To investigate the added value of our intervention, we set up an experimental study, in which we compared students receiving the intervention with students who did not receive the intervention. In both conditions, students received feedback on the competencies they used within the PBL setting (e.g. collaboration and self-regulation), and got the opportunity to use this feedback for pursuing personal learning goals. Only in the experimental condition however, students were actively stimulated and personally supported to go through a self-developmental cycle: starting with personal goals, pursuing those goals in learning situations and reflecting on the experience.

Our main research question was: What is the effect of an additional self-development intervention on the development of SDL of first-year students within a PBL-program? We expected that the pedagogical control in the form of a self-development intervention would benefit first-year students in the development of their SDL (Long, 1989). Thus, we expected students in the experimental condition who received a self-development intervention on top of the PBL-program to report a larger increase in SDL, compared to students in the control condition who only received the PBL-program.

Materials and method

Participants

To test our research question, we conducted an experimental intervention study with first-year Bachelor Psychology students at the Erasmus University Rotterdam, over the span of one academic year. All 754 Dutch-speaking first-year students received an information letter and consent form during the first weeks of the study and were asked to participate in the research project. Students who were willing to participate and gave their consent before the second course of the curriculum were included. The initial sample consisted of 408 participants. Participants were aged 16 to 55 (M = 20.29, SD = 3.52). Table 1 shows other demographic properties of all students in the sample.

Table 1. Demographic characteristics of the total sample.

	Total	Experimental group	Control group
	N (%)	n (%)	n (%)	X^2 (df)	p
Sex Female Male	309 (75.7) 99 (24.3)	89 (74.2) 31 (25.8)	220 (76.4) 68 (23.6)	0.23 (1)	.633
Previous education Pre-university Higher vocational Other	242 (59.3) 148 (36.3) 18 (4.4)	63 (52.5) 48 (40.0) 9 (7.5)	179 (62.2) 100 (34.7) 9 (3.1)	5.66 (2)	.059
Nationality Dutch Non-Dutch	399 (97.8) 9 (2.2)	115 (95.8) 5 (4.2)	284 (98.6) 4 (1.4)	3.03 (1)	.082

The 408 students were randomly assigned to the experimental or the control group. Because of limited resources only 120 students were included in the experimental group. There were no differences between both groups on sex, highest previous education, and nationality at the start of the study (see Table 1). An additional independent samples t-test also showed no differences in age between the experimental (M = 20.48, SD = 3.43) and control group (M = 20.21, SD = 3.55), t(406) = 0.71, p = .478.

Design & procedure

Figure 1 gives an overview of the study design. During the first 5 weeks (pre-test) and the last 5 weeks of the academic year (post-test), all participants received questionnaires measuring self-directed learning and self-regulated learning (Collaboration was also assessed but will be not reported about in this paper). To facilitate self-reflection, all students received a personal outcome report on the questionnaires they completed. The reported feedback was related to students’ general responses to the subscales of the questionnaire. For instance, for Self-Management (subscale of the self-directed learning questionnaire), the feedback could look like this: ‘Generally, you strongly agree that self-management is characteristic of yourself’.

Figure 1. Study design, including educational components and measurement moments.

Control condition (PBL program)

During the academic year, students received the regular PBL-program consisting of eight 5-week courses. Each course always contained nine PBL-meetings of groups of 12 students and a tutor. Once every course, students received written feedback from their tutor about the professional behaviour and skills the tutor had observed during the PBL-meetings. As part of this feedback, a grade was given, based on a list of 15 statements about their professional behaviour and skills that were scored on a scale from 1 to 5.

All feedback provided students with information and the opportunity to independently form personal goals regarding three professional competencies: 1) Self-development and reflection, 2) Personal effectiveness and 3) Collaboration. These three competencies had been picked by teachers from the program as the most important professional competences students need to attain during their studies, and had been formulated by the same teachers in line with the overall educational aims of the program. Self-development and reflection referred to the ability to independently set learning goals, to recognize and create opportunities for learning, and to ask for feedback for improvement. Personal effectiveness referred to the awareness of strategies used and their effects. It also involved the ability to verbalize strategies and adapt them within an organizational context. Collaboration was defined as the ability to build relationships with other people, to communicate in an empathising way, to be open about your own vulnerabilities, and to be aware of other people’s roles, of the organization, and of the world around you. To ensure students understood the purpose of working with personal goals well enough, students got information about development of professional competencies at the start of the academic year, including basic information on the three professional competencies consisting of the name of the competency and a general description.

Students could practice the three competencies during the academic year. Practicing could be done through study activities like preparing and discussing the problem-cases and the literature in the PBL-meetings. Feedback could be asked for during the PBL-meetings through short peer- and self-reflections, but only if a student had made clear arrangements with the tutor in advance.

Experimental condition (PBL- program plus self-development intervention)

Students in the experimental group received the same regular PBL-program as students in the control condition, but additionally received a self-development intervention starting in the second course period. The self-development intervention gave students specific guidance on how to improve their professional competencies Self-development and reflection, Personal effectiveness, and Collaboration, including three individual consultations with a mentor. All mentors (who also acted as tutors in the program) were trained before the intervention started, and during the intervention they had intervision meetings with a mentor supervisor.

The intervention was presented to students within their online learning environment, with a welcome, a course overview and a set of assignments for the whole year. The intervention started with each student, two of their peers, their mentor and one of their parents filling out a feedback competency questionnaire, evaluating the student’s current level on the three professional competencies Self-development and reflection, Personal effectiveness and Collaboration. Subsequently, students had to reflect on the outcomes of the questionnaire, and define their most prominent learning goals for this academic year. Furthermore, the students were asked to write a personal development plan in the form of answering questions that guided them through the writing process. To help student to formulate learning situations in their personal development plan, students were provided with extensive online information on the three targeted competencies. Online, each competency had its own developmental path with an assignment library. Students could pick any assignment from the assignment library matching the level of their competency or think of a learning situation of their own.

Once students had finished their concept personal development plan (during the third course), they had their first individual mentor consultation to discuss and finalise their personal development plan with their mentor. Mentors stimulated and coached students to formulate specific, challenging, and attainable personal goals and to choose specific learning situations during the different courses of the first year. During the year, students gained experiences in different learning situations and reflected on these experiences, uploading their reflections in an e-portfolio. During the sixth course there was a second mentor consultation to monitor the students’ progress. During the seventh course, each student, the same two peers, parent and mentor completed the feedback competency questionnaire again, to reassess their level on the three competencies. The results were discussed with the students during the final mentor consultation in the eighth course.

Adjustments

Because the study was conducted during the academic year 2019–2020, we had to make some adjustments because of corona restrictions: For all students, the regular PBL-meetings were done online through video conferencing from the sixth to the eighth course. Also, the regular tutor feedback was turned into a pass or fail, so there were no separate scores on 15 statements anymore; For the intervention groups, the mentor consultations during the sixth course had to be postponed for about a week, and the two remaining mentor sessions were done online, through video conferencing.

Instruments

SDL was measured with two different scales, as we wanted to gain more insight into which aspects of SDL were stimulated. The Self-Directed Learning Readiness Scale for Nursing Education measures ‘the degree the individual possesses the attitudes, abilities and personality characteristics necessary for self-directed learning’ (SDLRS-NE; Fisher et al., 2001). It consists of three subscales with each 40 statements: Self-Management, Desire for Learning and Self-Control. Items are scored on a 5-point Likert scale (1 = strongly disagree, 5 = strongly agree). We used the total score as well as the three subscale scores. Self-Management is defined as the ability to organise one’s own learning process. An example statement is: ‘I manage my time well’. Desire for Learning is related to the personality characteristic of curiosity for new information and has statements like ‘I enjoy learning new information’. Self-Control is about the attitude toward one’s own learning process: e.g. ‘I prefer to set my own goals’. The Cronbach’s alpha for the total scale was .92 and for the subscales .86, .85 and .83 respectively, in a sample of 201 bachelor nursing students (Fisher et al., 2001). For this study the SDLRS-NE was translated to Dutch using a backward translation design.

The second measure for SDL was the adjusted Dutch version of the Self-Regulation of Learning Self-Report Scale (SRL-SRS; Lucieer et al., 2016). Self-regulated learning is often seen as a construct largely overlapping with SDL (Loyens et al., 2008). However, whereas the SDLRS-NE focuses more on students’ sense of control and initiative, the SRL-SRS focuses more on the actual ability to plan, monitor, reflect and adjust actions and motivations to attain personal goals. The scale consists of 40 items divided over six subscales: Planning, Self-Monitoring, Evaluation, Reflection, Effort and Self-Efficacy. Items are scored on a 4 or 5-point Likert scale, depending on the subscales. We used the scores of each of the six subscales. The subscale Planning measures how students approach a task in advance of execution; Self-monitoring is measuring how well students monitor their progress during task performance; Evaluation is about checking the process and outcomes after task execution; Reflection is about what students learned from task execution; Effort is about concentration and persistence; and Self-Efficacy measures students’ beliefs that they can successfully execute their tasks. Cronbach’s α for the subscales were .86, .79, .79, .96, .83 and .80 respectively, in a sample of 949 first- and third-year medical students (Lucieer et al., 2016).

Results

Dropout analysis

Figure 2 shows the participant flow during the study. Only 61.5% of the participants completed both the pre- and post-test. Dropouts were generally a bit older (M = 20.96, SD = 4.53) than completers (M = 19.88, SD = 2.62; t(221.75) = 2.71, p = .007). Males dropped out more frequently (47.5%) than females (35.6%; X²(1) = 4.47, p = .035). Students with other pre-university education (colloquium doctum and foreign education) dropped out more frequently (66.7%) than pre-university education (35.1%) and higher vocational education (40.5%; X²(2) = 7.46, p = .024). Non-Dutch students dropped out more frequently (88.9%) than Dutch students (37.3%; X²(1) = 9.88, p = .002). From all participants, 38.9% of the dropout was due to discontinuation of the bachelor, the remaining 61.1% dropped out because of other reasons.

Figure 2. Participant flow.

The experimental group had comparatively more dropout (48.2%) than the control group (34%), X² (1) = 8.20, p = .004. However, there was no difference in average age between dropouts in the experimental (M = 21.08, SD = 3.95) and control group (M = 20.88, SD = 4.87; t(155) = .277, p = .783). Also, there were no differences in sex, highest previous education and nationality (see Table 2). At the pretest, dropouts in the experimental and control group did not differ significantly on (any of the subscales of) the SDLRS-NE and SRL-SRS (Table 3). Finally, students in both groups had comparable reasons for dropout: In the experimental group 39.0% dropped out because of discontinuation of the bachelor, and 61.0% because of unknown reasons, whereas in the control group these numbers were 38.8% and 61.2%.

Table 2. Demographic characteristics for the dropouts in both groups.

	Experimental group	Control group
	n (%)	n (%)	X^2 (df)	p
Sex			0.36 (1)	.550
Female	43 (72.9)	67 (68.4)
Male	16 (27.1)	31 (31.6)
Previous education			5.46 (2)	.065
Pre-university	25 (42.4)	60 (61.2)
Higher vocational	29 (49.2)	31 (31.6)
Other	5 (8.5)	7 (7.1)
Nationality			0.55 (1)	.457
Dutch	55 (93.2)	94 (95.9)
Non-Dutch	4 (6.8)	4 (4.1)

Table 3. Average scores on SDLRS-NE and SRL-SRS scales for dropouts in both groups.

	Experimental group	Control group
	M (SD)	M (SD)	t (139)	p
SDLRS-NE	152.37 (13.90)	151.74 (13.38)	0.26	.792
Self-management	45.61 (6.51)	45.39 (5.99)	0.20	.840
Desire for learning	49.12 (3.88)	48.28 (4.61)	1.10	.274
Self-control	57.65 (6.13)	58.08 (5.85)	−0.41	.680
SRL-SRS
Planning	15.98 (3.46)	15.87 (3.18)	0.20	.844
Monitoring	17.78 (2.90)	17.11 (2.89)	1.33	.186
Evaluation	29.76 (4.01)	29.96 (3.32)	−0.30	.762
Reflection	19.96 (4.53)	21.02 (3.06)	−1.66	.100
Effort	23.06 (3.57)	22.52 (3.94)	0.80	.423
Self-efficacy	20.94 (3.31)	20.21 (3.85)	1.14	.258

For the remaining 251 participants that completed both pre- and post-test, we found no differences between the experimental and control group on all scales of the SDLRS-NE and SRL-SRS of the pre-test, except for the SRL-SRS subscale Monitoring. At pre-test the control group scored slightly higher (M = 17.79) on Monitoring compared to the experimental group (M = 16.69; t(249)= −2.30, p = .022).

Main analysis

To test whether self-directed learning increased more in the experimental group than in the control group, we conducted 2 × 2 mixed ANOVAs with group (experimental and control) and time of measurement (start and end of academic year) as independent variables, and the scores on the (subscales of the) SDLRS-NE and the SRL-SRS as dependent variables.

For the SDLRS-NE, the results are presented in Table 4. On the total score we found a main effect for time of measurement. More importantly, we found a significant interaction effect, with the experimental group (+2.8%) showing a larger increase in score from the start to the end of academic year compared to the control group (+0.1%). Looking at the subscales, we didn’t find any significant effects on Self-Management and Desire for Learning. On Self-Control however, we found a main effect for time of measurement, qualified by a significant interaction, with the experimental group (+3.9%) showing a larger increase from the start to the end of the academic year compared to the control group (+0.6%).

Table 4. Average scores on pre- and post-test SDLRS-NE scales, and outcomes ANOVAs.

	Pre		Post		ANOVA
Scale	Exp. M (SD)	Control M (SD)	Exp. M (SD)	Control M (SD)	Time of measurement	Group	Time x Group
Total	152.82 (14.09)	155.26 (13.75)	157.02 (13.30)	155.46 (14.36)	F(1,249) = 7.16, p = .008, η^2partial = .028	F(1,249) = .06, p = .816, η2partial < .001	F(1,249) = 5.92, p = .016, η2partial = .023
Self-management	46.62 (6.32)	47.81 (6.40)	47.97 (6.93)	48.02 (7.23)	F(1,249) = 3.87, p = .050, η^2partial = .015	F(1,249) = .46, p = .499, η2partial = .002	F(1,249) = 2.06, p = .153, η2partial = .008
Desire for learning	48.23 (4.76)	48.93 (4.87)	48.80 (4.14)	48.56 (4.55)	F(1,249) = 0.12, p = .730, η^2partial < .001	F(1,249) = .14, p = .709, η2partial = .001	F(1,249) = 2.52, p = .114, η2partial = .010
Self-control	57.97 (5.77)	58.52 (5.48)	6.25 (5.00)	58.87 (5.26)	F(1,249) = 14.64, p < .001, η^2partial = .056	F(1,249) = .33, p = .564, η2partial = .001	F(1,249) = 7.77, p = .006, η2partial = .030

For the subscales of the SRL-SRS, the results are presented in Table 5. For Planning we only found a main effect of time of measurement. Planning increased from the start to the end of the academic year for both groups. For Monitoring and Evaluating we found a main effect for time of measurement, and a significant interaction effect of time of measurement and group. Students in the experimental group reported a larger increase over time on both the Monitoring (+8.0%) and the Evaluating (+6.9%) subscale, compared to the control group (+1.7% and + 2.6%, respectively). On Effort, no main effects were found, but only a significant interaction effect. Whereas the Effort score increased in the experimental group (+1.1%), it decreased in the control group (−3.2%). No significant effects were found on the subscales Reflection and Self Efficacy.

Table 5. Average scores on pre- and post-test SRL-SRS scales, and outcomes ANOVAs.

	Pre		Post		ANOVA
Scale	Exp. M (SD)	Control M (SD)	Exp. M (SD)	Control M (SD)	Time of measurement	Group	Interaction
Planning	14.67 (4.03)	15.69 (3.69)	16.28 (3.33)	16.55 (3.53)	F(1,249) = 24.07, p < .001, η^2partial = .088	F(1,249) = 1.86, p = .174, η2partial = .007	F(1,249) = 2.22, p = .137, η^2partial = .009
Monitoring	16.69 (3.43)	17.79 (3.21)	18.02 (2.91)	18.09 (2.95)	F(1,249) = 13.56, p < .001, η^2partial = .052	F(1,249) = 2.16, p = .143, η2partial = .009	F(1,249) = 5.50, p = .020, η^2partial = .022
Evaluating	29.51 (3.44)	30.52 (4.07)	31.54 (3.81)	31.31 (4.16)	F(1,249) = 23.36, p < .001, η^2partial = .086	F(1,249) = .58, p = .448, η2partial = .002	F(1,249) = 4.48, p = .035, η^2partial = .018
Reflection	19.74 (4.46)	20.49 (4.03)	19.56 (5.09)	19.97 (4.45)	F(1,249) = 0.65, p = .422, η^2partial = .003	F(1,249) = 1.50, p = .222, η2partial = .006	F(1,249) = 0.15, p = .696, η^2partial = .001
Effort	23.90 (3.58)	24.11 (3.75)	24.15 (3.62)	23.33 (3.93)	F(1,249) = 1.06, p = .304, η^2partial = .004	F(1,249) = .39, p = .532, η2partial = .002	F(1,249) = 3.93, p = .049, η^2partial = .016
Self-efficacy	2.41 (3.25)	19.94 (3.28)	2.90 (3.20)	19.76 (3.39)	F(1,249) = 0.46, p = .498, η^2partial = .002	F(1,249) = 3.54, p = .061, η2partial = .014	F(1,249) = 2.01, p = .157, η^2partial = .008

Subgroup analysis

To explore whether some groups had profited more from the intervention than others, we split our experimental group in 5 subgroups based on their initial SDLRS-NE total score at the beginning of the year.

A one-factor ANOVA showed a significant difference in change from pre- to post-test between subgroups (F(4,56) = 3.97, p = .007, ƞ²_partial = .221). Figure 3 shows that students in the experimental group with a lower initial SDLRS-NE score had a large increase in self-directed learning readiness score during the year, whereas students with a high initial SDLRS-NE score remained at roughly the same level.

Figure 3. Mean change from pre- to post-test in total SDLRS-NE-score (as a percentage) for subgroups in the experimental condition based on their total SDLRS-NE-score at pre-test. Error bars represent the 95% confidence intervals.

Discussion

This study investigates the effect of a one-year self-development intervention consisting of a year-long developmental cycle of setting personal goals, getting experience in learning situations, reflection, and discussion with a mentor, on the development of self-directed learning of first-year students within a PBL-based university program.

As hypothesized, we find that students receiving the intervention in addition to their PBL program report a bigger improvement than students only receiving the PBL program on several aspects of self-directed learning: self-control, the ability to monitor improvement during task performance, and the ability to evaluate process and outcome after the execution of a plan. Furthermore, receiving the intervention also seems to counter the drop in willingness to invest maximal study effort found in the group not receiving the intervention. We find these benefits even when implementing our intervention in a regular PBL program already meant to promote SDL, which is reflected in overall improvements on aspects like planning. So, we can conclude that our self-development intervention is effective for improving several aspects of self-directed learning for first-year students beyond the benefits of a PBL-approach itself.

The main aim of our study was to find out whether a self-development intervention promotes the aspects of SDL needed for self-development during lifelong learning. Whereas the self-development intervention from Colomer et al. (2021) only impacted students’ self-knowledge, we find benefits not only on aspects related to monitoring and evaluating study behaviour, but also on students’ attitude toward their own learning process, and the willingness to invest effort, which are important conditions for self-development. The intervention explicitly offered assignments and mentoring related to monitoring and evaluating, which may explain the benefits on these aspects. Also, during an evaluation discussion we had with the mentors after the intervention, they told us that students in the experimental group first saw the intervention as something that ‘had to be done’, but later in the year, due to improved reflection, they became much more motivated and convinced of the utility of developing their professional competencies. Almost all students also indicated they had appreciated the guidance provided by the mentor. So that might explain some of the improvements on attitude and effort. For some other aspects of SDL, the intervention seemed to be less effective. First, the intervention offered a planning assignment, but the mentors told us that students thought this assignment was difficult and needed mentoring to make it work. This may explain why we did not find any intervention benefits on planning. An alternative explanation might be that the PBL program already provided sufficient planning support, as both the experimental and control group improved on this aspect of SDL. Most unexpectedly, we didn’t find a benefit on self-reflection, although the mentors were convinced that students had become better on this aspect, largely due to all the assignments the students had made. Perhaps students need to go through the intervention cycle more than once (like in the study by Colomer et al.), in order to find measurable benefits on self-reflection aspects of SDL. But overall, the results are very promising given the relatively modest scale and duration of our intervention.

Assuming that our first-year students are relatively low in SDL at the start of the academic year, our results also seem to be in line with the theory of Long (1989) that students low in SDL need more pedagogical control in developing SDL. In this study, all students received a regular PBL-based university program with feedback, including the opportunity to organise with the tutor to practice with personal goals during PBL-meetings. Students however still seemed to benefit from the extra pedagogical control that was offered in the intervention. Further support for the benefits of pedagogical control for low SDL learners comes from our exploratory analysis. This analysis shows that in the intervention group, students with the lowest self-directed learning scores at baseline seem to have the highest increase in SDL scores. Interpretation of these exploratory results should of course be done cautiously, as the student subgroups were relatively small, but as a whole our results clearly support the idea that students low in SDL benefit most from the pedagogical control provided in a self-development intervention.

Of course, this study also has some limitations that need to be considered when weighing our conclusions.

Firstly, although the overall level of drop-out (38.5%) is not exceptionally high for a one-year intervention in a higher education setting where students regularly quit or switch studies, there was relatively more drop-out in the intervention group than in the group without intervention. Fortunately, our pre-test results indicate that the remaining groups were still comparable in terms of starting levels of SDL (except for one subscale being in favour of the control group). However, it is still possible that the drop-out in the intervention group was larger, because students were less motivated to continue with the ‘extra’ intervention program. This may have resulted in an intervention group that was relatively more willing to invest in the development of their SDL. Although we do not have any indication that this was the case in our study, it cannot be ruled out as an alternative explanation for (part of our) results. So in a next study, we recommend to also provide the control group with activities that require a comparable time investment as the actual intervention, to make sure ‘motivation to participate’ can be ruled out as explanation for any effect.

Secondly, although we trained our mentors before the intervention, and followed-up with intervision meetings during the intervention, the mentor supervisor noticed from the conversations during these meetings that mentors sometimes took rather different approaches in their roles. Although this was partly intentional, as we instructed mentors to adapt to the individual needs of students, we do not know for certain whether each mentor was equally successful in providing the most appropriate support. So although our results do show that the overall intervention was successful in stimulating SDL, the exact role of mentoring in achieving this success is less clear. Follow-up research could try to monitor the actual activities of mentors more closely, in order to get a better understanding of the effective ingredients in the mentoring part of the process.

Thirdly, our intervention was hit by the global Covid outbreak from 2020. This impacted both the educational program and our intervention, although only during the final three courses of the year. We were able to mitigate most practical challenges by quickly switching to an online format, leading to only a minor delay in implementing the final mentoring sessions. Moreover, from the literature it is known that self-regulatory learning can still effectively be improved in an online setting, especially when it entails some form of teacher-student interaction (Eggers et al., 2021). In a recent study, Wong and Kan (2022) even implemented an online version of a group-based PBL-intervention during Covid times, and found improvements on both PBL abilities as well as on self-directed learning as measured with the SDLR (the same measure we used in our study). So the switch from offline to online education due to Covid might not necessarily have impacted the effectiveness of our intervention study. But of course, the crisis also had a more general impact on students’ well-being and their attitude towards their study, which may have affected our results. From student and mentor evaluations we also know that in the intervention group, students often stimulated each other on their personal goals during PBL-meetings, and that switching to online meetings due to Covid seemed to have stopped this group stimulation. So, it would be a very good idea to replicate our intervention under more ideal study conditions, to see whether the effect of our intervention is similar or even bigger than we found under the current circumstances.

In sum, this study provides a good starting point for both practice and research on the development of SDL in higher education. The experimental set-up allows for a relatively strong conclusion that providing a developmental intervention on top of a regular university-program stimulates several aspects of SDL for self-development in students. Of course, we are aware of the limitations of the current study, as we only applied our intervention within one year of one specific PBL-program of first-year psychology, with limited knowledge of the specific mechanisms at work within the intervention, and not yet finding effects on all relevant aspects of SDL like reflection. For example, when implementing our intervention in a non-PBL setting, additional activities may be needed to address those aspects of SDL that PBL is known to stimulate, in order to get comparable benefits from our intervention. Some things about the intervention may also need adaptation. For example, our choice for individual mentoring may not be affordable in many situations, so it could be useful to investigate whether group mentoring is also workable as part of the cycle. And it would also be very interesting to follow-up on the development of SDL-competencies in students over multiple years, to see how lasting the effects are. So further research into interventions to stimulate SDL needed for self-development is necessary, preferably into a broad variety of settings, using an experimental approach comparable to the current study, to allow for strong conclusions about effectiveness, but also including process measures, to get more insight in the active ingredients of each intervention. This way, we will be better able to optimally prepare students in higher education for self-development during their professional career.

Ethical approval statement

For this intervention study, ethical approval has been obtained from the Ethics review Committee of the Department of Psychology, Education, and Child Studies, Erasmus University Rotterdam (application 20–046).

Acknowledgments

The development of the intervention was supported financially by the Community for Learning and Innovation from the Erasmus University Rotterdam. We would like to thank the Department of Psychology, Education & Child Studies from the Erasmus University Rotterdam for providing extra time and resources to the first two authors, both full-time academic teachers, to create the intervention and to set-up and execute the intervention study. We would also like to thank Manon Stuit for setting up the mentor training and supervising the mentoring process.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Notes on contributors

Cynthia Elderson-Van Duin

Cynthia Elderson-Van Duin is working as an academic teacher and a teacher supervisor at the Department of Psychology, Education & Child Studies from the Erasmus University Rotterdam. She is an innovation specialist as well as a Clinical Psychologist. Currently she is working on her PhD aimed at improving eHealthcare in a health literacy diverse society. Her research interests are innovations that support people in maintaining or developing their autonomy and responsibility in taking care of their life and health.

Peter Johan Renkema

Peter Johan Renkema was also working as an academic teacher at the Department of Psychology, Education & Child Studies from the Erasmus University Rotterdam during the period of the intervention study. Currently Pim is working as mathematics teacher at a secondary school in the Netherlands.

Huib Karel Tabbers

Huib Tabbers is working as Assistant Professor of Educational Psychology at the Department of Psychology, Education & Child Studies from the Erasmus University Rotterdam. His research interests are online learning, cognitive strategies, metacognition, text comprehension and instructional design. The aim of his research is to develop practically relevant guidelines for the design of instructional materials for primary, secondary and higher education, in collaboration with schools and instructional design partners.