Recruitment.

PPTs were recruited between January 2020 and June 2021 in Belgium and the Netherlands, using a convenience sampling strategy [23]. A flyer was distributed within the network for PT clinical internships of Hasselt University, within two regional networks of PPTs in the southern Netherlands (Limburg and Den Bosch networks), and at several educational activities for physical and occupational therapists (e.g. symposia) in both countries. In order to collect a wide range of PPTs’ perspectives, a heterogeneous sample matching the following criteria was required [23]: (1) PPTs with different backgrounds in terms of work settings (e.g. private physical therapy practice, and rehabilitation centre); and (2) variation in years of experience in treating children with DCD. The PPTs supplied their demographic characteristics (age, work setting, graduation year, and years of experience in treating children with DCD) by completing a short questionnaire.

Individual interviews.

Individual interviews were conducted to gain insight into the individual reasons of PPTs about their choices in MLSs used to teach motor tasks to children with DCD [21, 24]. The framework described by Kallio et al. [25] was used to develop the interview guide. A preliminary semi-structured interview guide was developed by the authors who had clinical, educational, and research expertise in both motor learning and children with DCD. The interview guide started with introductory questions to get acquainted and to elicit information about PPTs’ experiences in treating children with DCD. Subsequently, more specific questions explored therapists’ use of MLSs with these children (Table 1). The interview guide contained suggestions for the interviewer for open-ended follow-up questions, prompts and probes which the interviewer could use to elaborate initial answers [25]. The preliminary interview guide was tested with pilot interviews to assess coverage and relevance of content, and to identify possible needs for reformulating questions and optimising the interview procedure [25]. After three pilot interviews with members of the target population, the interview guide was finalised (S2 File). The data from the pilot interviews were discarded.

Download:

• PNG
  larger image
• TIFF
  original image

Table 1. Main topics of the individual and focus-group interview guides.

https://doi.org/10.1371/journal.pone.0297119.t001

The first author (IvdV) and four master’s students conducted the interviews. Six students each received 35 hours of education to make them familiar with the interview guide and procedure, and to teach them in interview skills such as using prompts and probes. Education included: reading literature about interviewing, and about the topics motor learning and DCD; listening to and discussing the pilot interviews; and performing two interviews by themselves on which they received extensive feedback. Two of the six students experienced difficulties in mastering the interview skills, leaving four to conduct the interviews. The interviewers were encouraged to use the interview guide flexibly to maintain the flow of the interview [25]. Because previous interviews showed that it was difficult for PTs to express exactly what their ideas were regarding their use of MLSs in a specific situation with a particular child [8, 10], the PPTs recorded one treatment session in which they taught motor skills to a child (aged 5–12 years) with (probable) DCD. Preferably, the child was diagnosed with DCD. However, because the mean age of receiving a diagnosis of DCD in the Netherlands is 7.02 years (SD 1.79), and the process of diagnosis takes an average of 2.79 years (SD 2.13 years) [26], PPTs were able to video-tape a treatment session of a child with probable DCD if the child met all four DSM-5 criteria for DCD. Interviewers watched the videos in preparation for the interviews and, during these, referred to situations observed to encourage therapists to elaborate on their thoughts and choices.

The audiotaped interviews lasted approximately one hour. The interviews were held in the PPTs’ own workplace so that they could support their answers with demonstrations. However, due to Covid-19 restrictions, five of 12 individual interviews were conducted online with Skype or Google Meet [27].

Focus-group interviews.

Two focus-group interviews were planned: the first one after 10 individual interviews, and the second after the final individual interview. None of the participating PPTs had participated in the individual interviews. The focus groups were conducted in addition to the individual interviews to enhance data richness [21, 24]. The focus-group interviews: (1) deepened topics mentioned in the individual interviews; (2) clarified and elaborated on different points of view about the use of MLSs; and/or (3) determined that insights obtained from the individual interviews were shared by a larger group of therapists [21, 24]. The findings of the interview analyses prior to the focus-group interviews determined the main topics of these (Table 1). The topics of Focus group 1 emerged from the analyses of the first 10 individual interviews. The topics of Focus group 2 were modified after analysing Focus group 1 and more individual interviews. The structure of the focus-group interview guides was similar to that of the individual one: (1) an introductory question to get acquainted with each other; and (2) specific questions addressing the main topics. The focus-group interview guides (S3 File) were discussed and fine-tuned within the research team (IvdV, KK, ER, CB) to ensure the relevance and completeness of their content [21].

To moderate group discussions, the interviewer (IvdV) asked follow-up questions, used prompts and probes, and invited participants to share their thoughts. Furthermore, an assistant made notes, managed time, and ensured that all topics were discussed [21]. Each focus group had 6 to 10 participants, and was organized in a venue proposed by the participants [21]. The audiotaped focus-group interviews lasted approximately two hours.

Theme 2: Therapists’ teaching style.

This theme consisted of three categories: (1) indirect/direct style; (2) PPT characteristics; and (3) child characteristics (Table 3).

In general, two types of teaching styles could be recognized: indirect and direct styles. A greater part of the PPTs preferred an indirect style, using questions and/or manipulations of task and environment to guide children with DCD to the correct movement solution. However, some of them preferred a direct style, instructing children exactly what to do. For instance, one PPT talked about how she used demonstrations with extensive verbal guidance to improve jumping skills:

“For example, I demonstrate jumping while also guiding very verbally.” “I give that guidance, so that he takes over from me, and starts guiding himself, first out loud and eventually in his head.”

Reasons for preferring an indirect or direct style differed. Some were related to PPT characteristics: (1) it suited them because it is the style they prefer themselves or it matches their character; and (2) they had learned by experience or education that a certain style works best. For example, several PPTs argued that they preferred asking questions because they believed that children learn more when discovering their own movement solution:

“I prefer asking questions rather than giving feedback, because I think that it [the movement solution] sticks better when the child comes up with it itself.”

Other reasons were related to child characteristics. Some PPTs suggested that children needed to be of a certain age and stage of cognition to process the explicit instructions used in a direct style. Furthermore, all PPTs agreed that children with DCD experience problems in executive functions (e.g. motor planning, finding movement solutions, and reflecting on their own actions). For some of them, that is why they use an indirect teaching style, so that such children learn these cognitive skills by themselves. However, for others, this justified a direct teaching style, because they believed these children to be insufficiently capable of learning these cognitive skills:

“I do not use that [guiding with questions] in children with DCD, I do it with my other children, but not with them because they cannot reflect on their motor disabilities.”

Theme 3: Motivation.

This theme consisted of three categories: (1) motivation as a prerequisite for learning; (2) child characteristics; and (3) strategies to improve motivation (Table 3).

All PPTs explained how a child’s motivation guided their use of MLSs. A general assumption was that demotivated children will learn less. Several PPTs stressed the importance of motivation as a prerequisite for learning:

“Because success makes happy and positive, and, with positive experiences, learning improves, right? That is [scientifically] demonstrated.”

Furthermore, PPTs talked about how various child characteristics impacted a child’s motivation according to their opinion. They underpinned the problems in automatization and transfer (Theme 4), and the lower levels of perceived competence of children with DCD as main reasons for these children not being motivated to practice, and getting frustrated when experiencing tasks as being too difficult:

“If it is really difficult, and it goes wrong every time, I don’t think they [children with DCD] will practice.”

Other reasons mentioned were movement anxiety and behavioural aspects like resistance when children were experiencing a bad day or were fatigued.

The PPTs suggested various strategies to improve children’s motivation (Table 4). For instance, one PPT talked about using themes to improve motivation:

“The boy had no motivation, because he was playing when he had to come to me. I asked what he was doing, and he told me he was making a marble run. So, we drew marble runs when practicing writing readiness skills.”

Download:

• PNG
  larger image
• TIFF
  original image

Table 4. Strategies to enhance children’s motivation.

https://doi.org/10.1371/journal.pone.0297119.t004

Theme 4: Optimal level of practice.

This theme consisted of two categories: (1) experiences of success and failure; and (2) learning stages (Table 3).

The majority of the PPTs emphasized that the optimal practice level is when children are challenged, but most trials are still successful. They argued that if tasks are too easy or too difficult, children will not learn and will become demotivated. They talked about the relevance of experiences of success and failure. Most PPTs underlined the importance of success in children with DCD:

“You make sure that the child still can perform the exercises, and that the challenge is there. But I think that, in these children, it is even more important that they get positive experiences.”

However, several PPTs also talked about how errors enhanced learning:

“Sometimes, you have to do something wrong to know how you should actually do it.”

One of the child characteristics frequently considered when estimating the optimal practice level was the child’s learning stage (e.g. cognitive stage). However, perspectives on the use of MLSs within learning stages differed. In the early stage of learning, some therapists said they used more explicit instructions and feedback, while others strongly preferred using manipulation of tasks and environment without instructions and feedback. In one focus group, PPTs discussed the use of variation in the early stage of learning. Some of them reduced variation during practice to accelerate learning, while others deliberately introduced variation because of the varying contexts found in daily life. In reaction to a PPT that elaborated on how she used various types of ball to stimulate a child’s anticipation abilities in throwing, another PPTs said:

”I practice the basics of the skill [throwing] to make a child familiar with it, and start varying in a later stage.”

As learning progressed, PPTs agreed more on increasing variation and using dual tasks to enhance transfer (Theme 4). However, some PPTs said they still used explicit instructions and feedback as well to optimize performance, while others did not use these in the later stages.

Theme 5: Automatization and transfer.

This theme consisted of three categories: (1) specific learning disabilities of children with DCD; (2) strategies to enhance automatization; and (3) strategies to enhance transfer (Table 3).

Most PPTs referred to the problems with automatization and transfer of skills as the specific learning disabilities of children with DCD. One PPT said:

“Automaticity takes much more time, so it is really important to give therapy in the best way in order to automatize [skills] as optimally as possible.”

The same PPT stressed that children with DCD have to keep practising tasks, as otherwise they forget how to perform them.

The PPTs suggested various strategies to enhance automatization. They stressed the importance of instructing parents and teachers to practise in daily life, and underpinned using the same wording in instructions and feedback:

“They [parents and teachers] should use the same wording as I do, because otherwise they [children with DCD] will never automatize.”

Furthermore, they suggested to practice tasks in similar ways throughout the various treatment sessions, and to decrease instructions and feedback when learning progresses to increase time for repetitions. They felt that with motivated children it was easier to achieve greater time on task.

PPTs also talked about their strategies to enhance transfer. For instance, they varied spatial and temporal constraints during practice (e.g. by continuously changing throwing direction and/or speed to improve the child’s catching abilities) to enhance anticipating to variable contexts in daily life:

”When they [children with DCD] know the movement pattern, than you should start changing to try to simulate other situations [from daily life].”

Other suggested strategies were: (1) simulating daily context by using dual tasks, or inviting other children to participate during treatment; (2) practising tasks that fit the child’s needs; and (3) using regular tools from children’s daily life (e.g. the child’s own bike).

Theme 6: Task complexity.

This theme consisted of four categories: (1) task constraints; (2) environmental demands; (3) children experience tasks as difficult to learn; and (4) PPTs experience tasks as difficult to teach (Table 3).

Most PPTs compared two types of tasks while elaborating on how tasks guide their use of MLSs. Frequently used examples were writing, cycling, rope skipping, and tying shoe laces. These tasks were compared to throwing, catching, running, and climbing. Because PPTs found it difficult to explain exactly how these tasks differed, this topic was extensively discussed in both focus groups. The overarching theme seemed to be task complexity, with four variables determining this identified: (1) task constraints; (2) environmental demands; (3) child and (4) therapist characteristics.

The PPTs mentioned following task constraints making tasks more complex: (1) multiple sequential steps; (2) dual tasking; (3) specific timing requirements; (4) bimanual coordination with both hands having different functions; and (5) the requirement to follow rules, for instance, in games. For instance, one PPT said:

“Eating is a bimanual skill, the hands must support each other, while doing opposite tasks” “I think that is what makes eating complex.”

They also pointed out that environmental demands could increase complexity, for example riding a bike in traffic is much more complex then riding a bike on an empty schoolyard:

“The child could ride a bike inside very well, but he refused to ride outside.” “Riding a bike depends on the person or the environment.”

Some PPTs noted that children may experience specific tasks as more difficult to learn:

“I find it [rope skipping] not difficult to teach, but I find it difficult to learn for the child [with DCD].”

Finally, some PPTs experienced tasks as more difficult to teach, which seemed to be related to their knowledge and experience:

“I find skipping [as locomotion] very difficult to teach to a child, probably one of the most difficult tasks.”

PPTs’ opinions on how tasks guided their use of MLSs varied. In both focus groups, they discussed how they used MLSs to improve performances of complex tasks (e.g. cycling or rope skipping). MLSs varied from explicit instructions, in which the child was told step by step how to ride a bicycle, to implicit strategies with manipulations of task and environment without using instructions and feedback, for example by letting the child ride the bicycle of a hill to increase speed.

PPT characteristics.

Most PPTs experienced difficulties putting their thoughts into words about which characteristics led their clinical decision making. They stressed that the interaction of child, task, and environment guided their choices, as suggested by the ’hybrid model of DCD’ [7]. However, the results of this study showed that PPTs choose MLSs intuitively, and that their clinical decision-making process was influenced by their own characteristics like knowledge, preferences, and beliefs as well. These findings are in line with a previous interview study that explored PTs’ perspectives on the construct of motor learning and their experiences of its implementation in clinical practice [10]. PTs stated that their use of MLSs was guided by intuition, and that their limited knowledge was an important barrier to implementation [10]. The importance of knowledge has also been demonstrated in several think-aloud studies investigating PTs’ clinical decision-making processes in rehabilitation, showing that knowledge from prior clinical experience, education, scientific research, and mentors or colleagues influenced their clinical decision making [8, 33, 34]. For optimal clinical decision making, PPTs require knowledge about: the use of MLSs to teach motor tasks (including adapting MLSs to child and task); the learning disabilities and associated problems of children with DCD; and basic knowledge about child development. The results of this study indicate the importance of the level of education on these topics. Specifically, PPTs’ knowledge about implicit and explicit motor learning approaches appeared limited. A need for more education has also been stressed by previous research [8, 10].

Child characteristics.

PPTs elaborated on how specific characteristics of a child guided their use of MLSs. Various child characteristics were identified. However, because of large variation in suggested characteristics and preferred MLSs, more research is required to gain insights into how the identified characteristics can guide PPTs’ decisions. One child characteristic that PPTs frequently mentioned guiding their choice of an indirect or direct teaching style was the presence of deficits in executive functions. As a result of deficits in inhibitory control, working memory, and attention, children with DCD have problems in planning and organizing activities of daily life [7, 12]. Based on their assumption whether executive functions could be trained or not, some PPTs preferred an indirect teaching style, while others preferred a direct style. Research shows that executive functions in children can be improved by training [35, 36]. Advancing critical and creative thinking, and problem-solving in movement situations encourages the development of executive functions [37]. Using questions that require children to think about movement solutions and then debriefing them about their actions is a frequently used strategy. Another is to place children in movement situations that challenge them to think about movement solutions [37]. Both strategies were used by the PPTs with indirect teaching styles. Because executive functions are important in many daily life activities (e.g. in learning at school, and in social interactions), and can be trained, PPTs can adopt an indirect teaching style to enhance the development of these executive functions when teaching motor tasks [37, 38]. PPTs choices based on the characteristics learning stage, the presence of learning disabilities, level of motivation, and level of perceived competence will be discussed in the section ‘Key elements in motor learning’.

Task complexity.

Task complexity was identified as the most important task characteristic guiding PPTs’ choice for MLSs. It appeared a complex construct described by four variables: (1) task constraints; (2) environmental demands; (3) child characteristics; and (4) therapist characteristics. The ‘challenge point framework’ conceptualizes complexity as a result of the combination of child, task and environment to which the PPTs should adapt their MLSs, which is in line with the results of our study [39]. The framework distinguishes two types of difficulties: the nominal task difficulty is defined by the task constraints, and is considered to reflect a constant amount of difficulty; the functional task difficulty is determined by the experiences of the individual (e.g. novices experience tasks as more difficult than individuals who have already performed those tasks) and environment (e.g. throwing outside in windy circumstances is more challenging than throwing indoors) [39]. This study also demonstrated that PPTs can experience teaching specific motor tasks as more or less difficult based on their knowledge and experiences.

PPTs’ opinions on which MLSs to use in complex tasks (e.g. riding a bike, tying shoe laces, and writing) differed: some used specific instructions focusing on the planning of these motor tasks, while others chose to provoke the correct movements by manipulating task and context. According to the international DCD recommendations, evidence-based methods like CO-OP and NTT can be used to teach motor tasks to children with DCD: CO-OP focuses mostly on motor planning, while NTT focuses on manipulating task and context [16–18]. Some therapists explained that they chose specific MLSs because they were trained in CO-OP or NTT. However, other reasons for choosing to focus on motor planning in complex tasks were given as well, demonstrating that PPTs own characteristics influenced their choices: (1) PPTs did not know how to manipulate complex tasks and its context; (2) it suited their own preference in learning; and (3) they believed that children needed to learn motor planning to advance learning in daily life.