Original Research

Reducing youth unemployment beyond the Youth Wage Subsidy: A study of Simtech apprentices

Mogantheran Naidoo, Muhammad E. Hoque

Received: 28 July 2016; Accepted: 17 Oct. 2016; Published: 24 Mar. 2017

Copyright: © 2017. The Author(s). Licensee: AOSIS.
This is an Open Access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Abstract

Orientation: South Africa currently has the twin challenges of worsening youth unemployment and scarce skills that threaten its economic and social stability. Artisanal trades are an occupation category that strongly reflects this current problem. Simtech Training Institute in Durban, the study setting, currently trains artisan apprentices and facilitates their internship work placements.

Research purpose: The objective of this study was to identify some of the critical success factors that differentiated Simtech artisan apprentices who obtained permanent employment, compared to those who are currently unemployed.

Motivation for the study: The main motivation of the study was to improve the conversion rate of artisan apprentices to permanently employed artisans.

Research design, approach and method: The study was a cross-sectional study conducted among 51 artisan apprentices who had graduated over the past 3 years at Simtech and who were selected randomly. An online questionnaire comprising primarily Likert scale type questions was utilised to obtain the responses from the sample. Factor analysis was used to remove scale items from the independent variables that did not impact the variability sufficiently. Then the remaining scale items that impacted variability significantly were combined and categorised as new composite independent variables. Logistic regression analysis identified success factors for permanent employment of Simtech graduates.

Main findings: Internship or workplace environment had a statistically significant impact on permanent employment. Youth work ethic had a minor impact on permanent employment status – albeit not a statistically significant one.

Practical/managerial implications: These findings showed that improving the internship/workplace environment can reduce youth unemployment and address skills scarcity.

Contribution: Internship host companies and other stakeholders need to urgently focus on improving the quality of the internship/workplace environment experienced by artisan apprentices rather than just on the intake number of artisan apprentices that the Youth Wage Subsidy has encouraged to date.

Introduction

At present, South Africa faces the twin challenges of rising youth unemployment rates and worsening scarce skills required to drive economic growth. Youth unemployment currently sits at 70% (StatsSA, 2014). This situation is likened to a ‘ticking time bomb’ (Burnett, 2014).

There is no quick fix for the current high youth unemployment levels (Yu, 2013). In this context, the Employment Incentive Bill was introduced on 01 January 2014 and is better known as the Youth Wage Subsidy. The objective of this legislation was to encourage the employment of young people with limited work experience, by using tax incentives (Rankin & Roberts, 2011).

South Africa currently has 30 000 fewer artisans than it requires based on vacancies (Akoojee, 2013). There is thus a major focus on artisan training schools and apprenticeships in the workplace to produce more artisans for the economy (Simkins, Rule & Bernstein, 2007). The effectiveness of such workplace apprenticeships is an important lever to increase qualified artisan output.

The youth unemployment problem becomes even more dire when looking at artisan trades. Current artisan apprenticeship programmes were not having the desired effect in reducing youth unemployment while producing scarce skills for the economy (Oseifuah, 2010). A lot more needed to be done to improve the output of such apprenticeship programmes (Smith, Jennings & Solanki, 2005).

This study sought to solve the problem of what critical success factors ensure that apprenticeship programmes yield the desired output of qualified artisans who are permanently employed and contributing to South Africa’s economy (Witte, Rothmann & Jackson, 2012).

Research objectives

The research objectives of the study are to:

• investigate whether the internship or workplace environment of Simtech’s young artisanal apprentices significantly impacted their attainment of permanent employment after the apprenticeship period
• identify whether the work ethic of Simtech’s young artisanal apprentices significantly impacted their attainment of permanent employment after the apprenticeship period
• determine whether the employability skills of Simtech’s young artisanal apprentices significantly impacted their attainment of permanent employment after the apprenticeship period.

Literature review

Youth unemployment

Youth unemployment has been one of the most concerning trends in the global labour market (Ogbuanya, 2015). It was particularly prevalent in developing economies like South Africa (Mlatsheni & Rospabé, 2002). High youth unemployment rates contributed to a number of socio-political problems in countries, such as high crime rates, high poverty levels and radical political upheavals – as in the Arab Spring uprisings that swept through north Africa and most notably Egypt a few years ago (Akoojee & Gonon, 2013). South Africa was starting to experience similar rhetoric with the emergence of the Economic Freedom Fighters political party that has sought to obtain economic freedom for its predominantly young support base (Burnett, 2014).

Developed countries such as Germany and Japan have not been immune to youth unemployment and skills shortages (Akoojee & Gonon, 2013). Declining population rates coupled with a disinterest by their young people in blue-collar artisanal trades has resulted in a ‘lost generation’ of artisans emerging in these countries (Wildschut, Meyer & Akoojee, 2015). This has meant that these countries needed to import such scarce skills into their economies, and this has driven a global scarcity of artisans (Horwitz, 2013).

Role of the apprenticeship system in South Africa

The apprenticeship system in South Africa had a turbulent past. During apartheid, it was used as a job reservation gatekeeper to prevent black people from accessing these trades (Smith et al., 2005). This is one of the reasons for the current shortage of artisans in South Africa. The majority of the population was prevented from accessing apprenticeships – which created a ‘lost generation’ of artisans in the South African labour market (Kruss & Wildschut, 2015).

Another important reason for the current shortage of artisans has been the discontinuation of many State-Owned Enterprise (SOE) apprenticeship schools after the advent of the new democracy in 1994 (Kruss et al., 2014). This was necessitated by the need for these organisations to become more profitable. As a result, artisan apprentice training was deemed a non-core activity and many training schools were shut down (Akoojee, 2013). This has caused a severe loss of training capacity in South Africa for artisanal trades. Many former instructors and trainers who were retrenched more than a decade ago are now no longer around. It would indeed take a lot to re-establish the institutional capacity of these SOE training schools (Wedekind, 2013).

Over the past 20 years, blue-collar occupations such as artisanal trades have become less and less popular – as most young people have sought university education to access occupations in the service industry and information technology sectors (Schlechter, Faught & Bussin, 2014; Wildschut et al., 2015). South Africa is not alone in this regard, as there was a similar trend in developing economy peers such as India and Nigeria (Ukachi, 2015). However, recent developments suggest that interest in artisanal trades could be increasing again. A phenomenon such as the widespread drought in South Africa has also popularised an artisanal trade such as plumbing – particularly in the rural areas of the country (Wildschut et al., 2015).

Private institutions such as the Simtech Training Institute – the focus of this research – have been established to increase the number of artisan apprentices graduating into the labour market. They have been trying to fill the gap caused by the closing down of SOE training schools. While organisations like Simtech provide most of the theoretical and practical aspects of a trade onsite at their training school premises, they are reliant on industry companies to provide the apprentices with much needed workplace experience (Allais, 2012; Simon, Pauline, & José-Luis, 2014).

Success factors of employed apprentices

Youth work ethic

While acknowledging the existence of circumstances beyond the control of the youth worker, there is much that a youth worker can do through hard work and effort to influence their progress during the internship (Becton, Walker, & Jones-Farmer, (2014). South Africa has one of the lowest work ethics in the world (De Armas, Grove, Hiatt, Taylor-Roy & Quezada, 2014). This is reflected in many studies – most notably ‘The World Economic Forum – Global Competitiveness Index’, where South Africa ranked 134 out of 144 countries that participated in the benchmarking exercise (Becton et al., 2014). Accusations of an ‘entitlement mentality’ creeping into the minds of young South Africans have been bandied about. Unrealistic dreams of large rewards with very little efforts have also played a part in diminishing the work ethic of South Africa’s young people (Rauscher, Wegman, Wooding, Davis & Junkin, 2013).

Society bore the responsibility for assisting unemployed youth with social grants in the income provision system. It is thought that this created an entitlement mentality in young people and has diminished their will to work (De Armas et al., 2014). Work ethic was determined by a number of different factors. There was some consensus on certain key elements that constituted the construct of a work ethic. Reliability, initiative and determination were important (Deal et al., 2013; Roberts 2013). Character was highlighted as being extremely important (Cohen, Panter, Turan, Morse & Kim, 2014; Kruss & Wildschut, 2015). Accountability was also put forward strongly (Rauscher et al., 2013).

Workplace or internship environment

This section looked at the youth unemployment problem from the perspective of the institution or company that took in youth workers and were responsible for training and moulding them into productive, competent individuals. These institutions or companies sometimes received tax incentives such as skills levies or the Youth Wage Subsidy in return for fulfilling this role.

Mentorship quality

The quality of mentorship was extremely important for ensuring the success of the internship period and nurturing young talent within an organisation. Many studies looked at the quality of mentorship and presented a variety of viewpoints (Doerwald, Scheibe & Van Yperen, 2015; Patil, 2015).

The initial chance taken on a young individual by an astute mentor who saw the potential of the individual, often unearthed gems who went on to be highly successful employees who contributed much to the success of organisations (Mulligan-Ferry & Nugent, 2016). Given the impact such decisions could have on both companies and young workers, it was important that mentors were chosen carefully. Unfortunately, far too often this selection of mentors was not done correctly (Patil, 2015). Either the mentor was overloaded with work and had insufficient time to fulfil his or her duties to the mentee or the mentor was disinterested in the growth of the youth worker and saw them as a threat to their position in the company. Indeed, in South Africa, with the introduction of the Youth Wage Subsidy, there was much resistance and antagonism from older workers towards young workers (Doerwald et al., 2015).

Training and development opportunities

It was of vital importance that relevant training and development opportunities are afforded to young workers during the internship period. This was to ensure that they are continuously learning valuable skills and improving as future employees (Du Toit & Roodt, 2008). According to Downs (2015), the investment in high-potential employees yielded significant results for companies in the long term. This view is supported by Boyle (2015), who stated that significant training investment had started to yield the desired outcomes – with more engineering interns becoming competent to meet the needs of industry in Scotland.

South Africa invested a large amount of money in training and development each year. This was largely collected via the Skills Levies Act and administered by Sector Education and Training Authority (SETA). The success of these organisations though has been questionable (Smith et al., 2005). More needs to be done to extract more value out of the training sessions relative to the amount spent on them. Ben-Hur, Jaworski and Gray (2015) discussed whether the training budgets were being spent on the right things.

Employability skills

ILO (2013) referred to core employability skills as those skills necessary for lifelong learning and they were transferable across career roles. ILO stated that these skills equipped young workers to adapt to continuous change in the world of work and prevented their skills from becoming obsolete. The ILO had narrowed down these core employability skills to learn, communication, teamwork and problem solving. Mourshed, Farrell and Barton (2012) stated that 43% of employers could not find the required employability skills in young workers. This showed that much more needed to be done to improve the employability skills of young workers – and thus enhance their employment chances.

In South Africa, there has been much focus on providing access to workplace internships via the Youth Wage Subsidy launched in 2014. This is a noble effort given that the first work experience plays a significant role in a young worker’s career. However, not enough focus has been put on the actual internship period, in preparing the youth worker for permanent employment jobs elsewhere or within the same company.

Research methodology

This study utilised a quantitative methods design. According to Creswell (2014), such a design was well suited to testing predetermined outcomes and statistically analysing significant correlations. The quantitative method was utilised by this study, so that attitude data could be gathered by instrument-based questions (Sekaran & Bougie, 2013).

The Simtech trade apprenticeship school was chosen as a suitable study setting, given the familiarity of the researcher with their operations and because of accessibility of their students who could be reached to participate in the empirical research. This institution nurtures and trains young people with technical and general workplace skills, over a certain period of time. These apprentices then serve time in the industry, working for numerous companies and gaining workplace experience. The Simtech institution and its apprentices was thus an ideal choice for carrying out this exploratory study on understanding their actual experiences during their internship period.

Simtech had 120 apprentices over the past 3 years who completed their training programme and workplace internship. These 120 graduates were the target population for the study and currently were either employed or unemployed.

A 95% confidence interval and 5% level of significance was used to calculate the required sample size for the research study. A sample size of 92 was needed to significantly represent the 120 Simtech apprentice graduates who have graduated in the past 3 years. A simple random sampling technique was used to select the samples for the study.

The data collection instrument was an online questionnaire. This was the most efficient method of collecting sufficient responses from a representative sample size of the target population within the desired timeframe (Creswell, 2014). The online questionnaire was designed to take only 15 minutes to complete – to encourage a higher participation and completion rate. Most respondents would not be English native speakers. Special care was taken to use simple English in the questionnaire and to make the questions as straightforward as possible. This was achieved by avoiding double-barrelled questions with ambiguous meaning.

Respondents without e-mail access were accommodated by the researcher administering the survey telephonically. The researcher then captured the data onto the eSurveyCreator system. It was anticipated that the data collection would take place over a 3- to 4-week period. The questionnaire was administered following a three-stage process. Firstly, there was an initial e-mail advising participants of the survey. Secondly, the actual link to the electronic survey followed 3 days later. Lastly, a follow-up e-mail was sent a week after the link was sent out – to thank those participants who had already responded and encouraging those participants who had not yet replied to do so.

A Likert scale was chosen to code the responses into quantifiable integer numbers ranging from 1 to 4. A four-point Likert scale was chosen deliberately – instead of the standard five-point Likert scale. The reason was to eliminate neutral responses, by forcing respondents to select negative or positive responses.

The online questionnaire was initially e-mailed to five people to test its functionality. Key aspects checked where the time needed to complete the survey, the clarity of the questions and the Cronbach’s alpha that verified the internal consistency of the items being tested. The Cronbach’s alpha was 0.91, which indicated that the data were reliable.

The study received an ethical clearance certificate from the ethics committee of the University of KwaZulu-Natal. Also, a permission letter was obtained from the CEO of Simtech Apprentice Training Institute, prior to questionnaires being sent out. Each respondent had the option of withdrawing from participation in the questionnaire at any time they wished, without any penalties. This was clearly explained in the letter of informed consent that accompanied the online questionnaire. The actual research instrument – the online survey – underwent strict ethical clearance scrutiny by the University of Kwa-Zulu Natal’s Research Ethical Clearance Committee. The field study only proceeded once such ethical clearance (Protocol reference number: HSS/0249/016M) was granted.

Results

Biographical information

Table 1 shows the summary of age distribution and internship period of the participants. About half the respondents (42%) were aged 28–30 years and the same number spent 1–2 years as an apprentice.

Ten statements comprised the construct of the youth work ethic. There was a noticeable difference in the answering choices of the employed and unemployed groupings. For example, 56% of the unemployed grouping felt that they should have a job, no matter how they performed – pointing towards an entitlement culture. It was found that 72% of the employed grouping was trusted by colleagues to do their work, while 68% of the employed grouping accepted responsibility for their mistakes (Figure 1).

FIGURE 1: Youth work ethic frequency distribution.

For the employed respondents’ grouping, the most frequent mode of answer on the Likert scale was ‘agree’. In contrast, for the unemployed respondents’ grouping, the most frequent mode of answer was ‘strongly disagree’ or ‘disagree’. Some 68% of the unemployed grouping strongly disagreed that their internship host company offered any apprentice a permanent job, while 60% further strongly disagreed that they were told they would get a permanent job at the end of the apprenticeship if they performed to a high standard. In stark contrast, the employed grouping largely agreed on every one of the ten items – pointing towards a largely positive internship or workplace environment (Figure 2).

FIGURE 2: Internship or workplace environment frequency distribution.

Most of the employed participants agreed to most of the statements regarding employability skills. In contrast, for the unemployed respondents grouping, the most frequent of answer was negative. About two-thirds (64%) of the unemployed grouping stated that they strongly disagreed that their internships equipped them with entrepreneurial skills (Figure 3).

FIGURE 3: Employability skills frequency distribution.

The Kaiser–Meyer–Olkin measure of sampling adequacy in Table 2 was greater than 0.5, which implied that the sampling adequacy was extremely good.

All scale items in Table 3 that had an Eigenvalue greater than 1 were retained in the model. An Eigenvalue equal to 1 was considered an average score. Table 3 illustrates the scale items that were subsequently included and excluded from the three composites that were then created: Composite_work ethic, Composite_workplace environment and Composite_employability skills. The remaining scale items that made up the three composite independent variables were then added to reach a sum total score of the various scale items and were transformed into binary using the following set rule. All cases that had a composite total score less than the mean score were transformed to a 0; and all cases that had a composite total score equal to or greater than the mean score were transformed to a 1. This transformation function prepared the three composite independent variables for the logistic regression analysis against the dichotomous dependent variable.

The Nagelkerke R-squared value was 0.653, which implied that 65.3% of the variability in the dependent variable could be predicted by the model. This implied that the model, formed with the three composite variables refined by the factor analysis was strong.

The classification shown in Table 4 indicated that the model correctly predicted the employment status of 86.3% of all cases captured by the data. This further reinforced the strength of the model.

The variables included in the equation in Table 5 were the composite workplace environment and the composite work ethic. The composite workplace environment was statistically significant at a value of 0.001 (p < 0.05). Although composite work ethic was not statistically significant at a value of 0.998 (p > 0.05), it still made it into the final model, indicating that it had some impact on the prediction model. An artisan apprentice who was happy with his or her workplace environment was 18.33 times more likely to be employed than an artisan who was not happy with his or her workplace environment.

Discussion

The biographical information showed that most respondents were in the 28- to 30-year bracket, implying that a level of maturity was needed to complete an apprenticeship properly and obtain permanent employment. Most respondents had an internship period of less than 2 years, which showed that companies were possibly just employing them to derive the benefit of the Youth Wage Subsidy tax benefit. This shorter duration could impact the skills transfer and depth of knowledge acquired by apprentices, to become qualified, competent artisans.

The main critical success factor was the importance of a positive and enabling internship or workplace environment in determining the permanent employment status of artisan apprentices. Artisan apprentices who enjoyed a positive workplace environment were a staggering 18.3 times more likely to obtain permanent employment compared to their unemployed peers (p = 0.001). The factors that comprise internship or workplace environment were under the direct control of internship host companies, and these companies can utilise this insight to improve their training programmes offered to artisan apprentices. The internship/workplace environment factors examined were mentorship, training and development opportunities and talent pipeline management. The findings are discussed in more detail below.

The quality of mentorship was extremely important for ensuring the success of the internship period and nurturing young talent within an organisation. Past studies on the quality of mentorship (e.g. Mulligan-Ferry & Nugent, 2016) stated that the initial chance taken on a young individual by an astute mentor who saw the potential of the individual often unearthed gems who went on to be highly successful employees who contributed much to the success of those organisations. Therefore, it is important that internship host companies select mentors wisely and also equip them to adequately perform their functions. Unfortunately, far too often, this selection of mentors is done incorrectly (Patil, 2015). Either the mentor was overloaded with work and had insufficient time to fulfil his or her duties, or the mentor was disinterested in the growth of the youth worker and saw them as a threat to his position in the company.

The findings also revealed the vital importance of relevant training and development opportunities afforded to young artisan apprentices during the internship period. These training opportunities ensured that young apprentices were continuously learning valuable skills and improving as future employees. According to Downs (2015), the investment in high-potential employees yielded significant results for companies in the long term. This view was supported by Boyle (2015), where significant training investment had started to yield desired outcomes – with more engineering interns becoming competent to meet the needs of industry in Scotland.

South Africa invests a large amount of money in training and development each year. This is largely collected via the Skills Levies Act and is administered by SETAs. The success of these organisations though has been questionable. More needs to be done to extract more value from training sessions relative to the amount spent on them (Du Toit & Roodt, 2008). Ben-Hur et al. (2015) postulate whether the training budgets were being spent on the right things.

Progressive companies view the internship process as a way of securing the scarce skills they require to drive their business ambitions (Edmond et al., 2007). According to Drew (2015), there was a concerted effort made to capture young talent and the brightest minds in the labour market. This was further enforced by Rabbi et al. (2015) that talent management was a significant source of competitive advantage. Alkhalaf, Zaballero, and Alzahmi (2015) reiterated that company strategy was built around their talent pipeline management. The survey results contradict this theoretical view and suggest that further investigation into artisan development to meet scarce skills demand is warranted.

The other critical success factor discovered was that the composite_work ethic variable had some impact on the dependent variable. The factors that comprise youth work ethic can be directly influenced by the youth themselves – to improve their own chances of success by displaying attributes of desirable employees to their internship host company. The youth work ethic factors that were investigated were reliability, initiative, determination, accountability and character.

More than 50% of unemployed artisan apprentices did not complete allocated tasks within deadlines, while 96% of employed artisan apprentices completed their allocated tasks within deadlines. Clearly, going forward, all artisan apprentices need to ensure that they complete all their allocated tasks within deadlines. Employers are much more likely to permanently employ artisan apprentices who display such attributes. Arriving at work punctually had less of an impact on attaining permanent employment. Most unemployed and employed artisan apprentices attended work on time.

The survey results back up this theoretical perspective by clearly showing that those apprentices who did not quit tasks that became difficult ended up with permanent employment more frequently. It is noteworthy that 100% of all employed respondents did not quit difficult tasks. Young artisan apprentices must not give up on difficult tasks, but should rather persevere. They should request assistance early on from colleagues and supervisors when experiencing difficulties. Ultimately, they need to want to make a success of their apprenticeships. If this inner determination is lacking, it will clearly adversely affect their employment prospects.

It was shown that initiative was another important attribute in contemporary workplaces (Roberts, 2013). Young apprentices have to be able to motivate themselves to tackle more tasks and drive their own personal growth (Deal et al., 2013). With most employers having lean structures, mentors do not have much time to spend with each artisan apprentice. Young apprentices clearly have to create and grab opportunities to learn more and take on additional tasks. Those apprentices who took a lot of initiative by voluntarily working additional hours derived more value from their apprenticeships and were ultimately more successful in obtaining permanent employment.

The present study found that young artisan apprentices who accepted responsibility for their mistakes were more likely to attain permanent employment. All (100%) employed artisan apprentices accepted responsibility, while only 60% of unemployed artisan apprentices did so. Young artisan apprentices must learn to take accountability for their actions. They need to discard the trend of entitlement and blaming others for their present status in society. It is only once youth discard this mentality that they can move forward as better future employees.

It was concluded that character and personal demeanour were important components of a strong work ethic (Van der Walt & De Klerk, 2014). These characteristics often come from an individual’s background and upbringing (Cohen et al., 2014). Some common character traits were honesty and trustworthiness, which were often seen as hallmarks of successful high-performing employees. Character was something built up over a lengthy period of time, and yet could be destroyed quickly through negative actions (Schreiner, 2015). Youth workers had to demonstrate their good character repetitively and consistently to earn themselves a good reputation during their artisanal apprenticeships.

Conclusion

The study established that of the three independent constructs of youth work ethic, internship or workplace environment and employability skills, only internship or workplace environment had a significant relationship with the dependent variable of permanent employment attainment. Youth work ethic had some influence, but was not significant. While such a correlation is not necessarily causal in nature, it does prove that a strong and supportive internship or workplace environment was an accurate predictor of success for young artisan apprentices attaining permanent employment. Conversely, should young artisan apprentices be placed in a weak and unsupportive workplace environment during their internship, their chances of attaining permanent employment were significantly diminished.

Recommendations

Based on the research findings, the following are recommended:

• Future artisan apprentices of Simtech should commit themselves to giving of their best throughout their internship period at their host companies and display a strong work ethic that makes them attractive to future employers.
• Internship host companies should urgently review the suitability and performance of the mentors they allocate to oversee the development of artisan apprentices. There must be some deliverables that mentors are accountable for.
• Government departments overseeing administration of the Youth Wage Subsidy and similar tax incentives should vet internship host companies more stringently to determine whether the contribution they make towards national artisan development outweighs the tax incentives they are benefiting from. Defaulters who are abusing the incentive programme must be taken to task and have appropriate penalties enforced to correct bad behaviour.
• The artisan apprentice curriculum should be modified to add more employability skills that are useful across a broader range of industries – to improve the employability of young artisan apprentices.
• The internship host company should view apprenticeships as a key part of their talent pipeline management system – to cultivate the scarce skills they need to grow.
• Young artisan apprentices should want to do the respective professions they enter apprenticeships for and should have a strong passion for that vocation. Apprentices should be taken from technikon or training colleges where they have had some technical exposure and post-schooling education. Currently, apprentices sourced straight from matric is not ideal, as many of these candidates lack the maturity to grasp the apprenticeship with the seriousness it deserves.

Limitations of the research study

The following limitations of the research study are acknowledged:

• This study focused only on Simtech artisan apprentices for the past 3 years. Simtech is one of many artisan training centres in South Africa. Hence, the results cannot be generalised to all artisan apprentices in South Africa.
• Most respondents in the survey do not speak English as their first language. Their understanding of the various questions was assumed, but there is the possibility of bias in their responses, because of a lack of understanding.

Acknowledgements

The authors declare that they have no financial or personal relationships that may have inappropriately influenced them in writing this article.

M.N. conceptualised the topic, carried out the data collection. M.N. and M.E.H. developed the methodology and analysed the data. M.E.H. wrote the manuscript. Both M.N. and M.E.H. approved the final draft of the manuscript.

References