Students’ Conceptions of 21st Century Education in Zimbabwe

The 21st century has ushered in accelerated new technologies, new knowledge and new information and the fourth industrial revolution (4IR) – causing rapid changes in the way humans live and work and resulting in an outcry for educational reforms globally. Many frameworks and debates centred on rethinking and reshaping the curriculum including what knowledge, skills, values and attitudes should be taught to prepare learners to thrive in the 21st century, have been developed. This study examined the students’ conceptions of 21st century education in four diverse schools in Zimbabwe. The study is framed by the powerful theory and the complexity theory. The study used a quantitative non-experimental correlation design to collect data from four secondary schools, three of which follow the ZIMSEC syllabus and the other, a private school that follows a Cambridge syllabus. The questionnaire was adapted from the Partnership for 21st century framework (P21) and (Ravitz, 2014) to fit the Zimbabwean context. A total of 236 students and 93 took part. Exploratory and confirmatory factor analyses and ANOVA were used to analyse the data. The results showed that the students’ conception of 21st century skills is defined by three factors: the learning of self-management, learning skills and values and learning process. Students most strongly endorsed self-management. It is envisaged that the proposed model of Zimbabwean students’ conceptions of 21st century skills will act as a baseline for theory, policymakers, head teachers, teachers and teacher education colleges to realign the current teaching to fit the 21st century curricula.


Introduction
Since 1980, the exponential growth in technology, the effects of the fourth industrial revolution, and globalisation's impact have necessitated changes in education and curriculum. Schools the world over are reshaping the way students learn to enable them to live and thrive in the 21st century (Williamson, 2013;Luterbach & Brown, 2011;Malik, 2018;Rizvi & Lingard, 2010;Joynes et al., 2019). Several authors have advocated for the need to equip learners with 21st century skills that will help them survive in a world marked by volatility, uncertainty, complexity and ambiguity (VUCA) (Gravett, 2019;Kennedy, 2019). This demand to equip learners with 21st century skills resonates in many frameworks globally and regionally (Joynes et al., 2019;Rizvi & Lingard, 2010) posit that although globalisation has transformed educational policy to mirror how the world has become a global village, nations, states, and regions need to reflect their local knowledge, values and attitudes in their educational systems. This view is supported by the differences between global and regional needs. Even within regional contexts/developing countries (referred to as Low and Medium Income Countries (LMICs), differences exist due to policy and labour market capabilities priorities. This also suggests the impact of globalisation and technological change is heterogeneous across contexts, and therefore, the future demand of the 21st century skills becomes context-specific (Joynes et al., 2019). For instance, the economy in LMICs are mainly informal, agro-based, and require little high-tech skills. In terms of education, most LCMIs are still trying to equip learners with basic literacy and numeracy skills. For these reasons, it becomes difficult for LMICs to leapfrog their educational reforms to address the need for knowledge economies before addressing basic literacy and numeracy skills. Joynes et al., (2019) mention some signs in some regional contexts where efforts to transition into knowledge economies are evident, but they are geographically and socially limited and only receive aid from private sectors. At the regional level, Singapore, South Africa and Kenya have embarked on curriculum changes as well, with Singapore promoting a new vision called Thinking Schools, Learning Nation (1997) (TSLN), aimed at fostering essential skills like critical thinking, character development in terms of discipline, motivation, keen sense of responsibility and sound moral values, and developing a culture of lifelong learning beyond the formal schooling years (Ministry of Education [MoE], 2005). Kenya also effected educational reforms to align education with skills in line with their Vision 2030, which espouses economic growth, provision of 21st century skills for a modern economy, essential knowledge, national values and social values that will enrich the Kenyan society (Otunga & Nyandusi, 2009;Care et al., 2017a as cited in Joynes et al., 2019). The South African education system promotes the teaching of 21st century skills and values needed and embedded within the curriculum. The curriculum aims to equip learners with higher-order skills such as problem-solving, creative and critical decision making, effective teamwork, communication and environmental literacy. However, the South African government has also prioritised inclusion, diversity, equity, and life in the 21st century. Regarding demands for 21st century skills, emphasis has been placed on citizenship and nationalism and soft skills such as confidence, self-motivation, and self-drive (Joynes et al., 2019).
In order to redress and align its curriculum to the demands of the 21st century, Zimbabwe, a largely agro-based country, also implemented a new curriculum, the Zimbabwe Curriculum Framework (ZCF) 2015-2022, in order to educate its learners to thrive in an environment that is growing more globalised and competitive (ZCF 2015(ZCF -2022. The ZCF is similar to popular 21st -century frameworks like the Partnership for 21st century framework (P21). The ZFC 2015-2022 outlined the same educational goals of preparing the 21st -century learner for the world of life and work. However, education is grounded in the philosophy of unhu/Ubuntu/vumunthu and local knowledge, values, and attitudes. Like Kenya, Zimbabwe's Ministry of Primary and Secondary Education (MOPSE) summed up the new curriculum as a tool that will move education from mere cramming of fact into an outcome-based realm where learnt knowledge contributes to the growth of society (MOPSE, 2014). This assessment is similar to the case of Kenya in the call for reform, skills and relevance. Nevertheless, as stated above, the challenges lie in how these skills are to be taught, teacher competency, school leadership, and policy development.
Therefore, it is essential to consider what knowledge, skills, values, and attitudes should be developed to succeed in life and work in the future. However, one of the challenges that arise when considering the knowledge, skills, values, and attitudes to equip 21st century learners lies in the socio-economic context that defines the demand for these skills. Joynes et al., (2019) argue that 21st century skills are heterogeneous in different contexts.
At a local level, Zimbabwe has also developed several policies to ensure the growth of the education sector. One of the major reports was the Presidential Inquiry into Education and Training (1999), also referred to as the Nziramasanga Report (1999) (MoPSE, 2018). However, Zimbabwe has taken a long time to implement changes to its academic-oriented education curriculum to incorporate the many recommendations made by the Nziramasanga Commission in 1999 (Gondo Maturure, Mutopa, Tokwe, Chirefu, Nyevedzanayi, 2019). Changes to the Zimbabwe educational framework finally commenced in 2017 with schools embracing the Zimbabwe Curriculum Framework, which was tailor-made for rapidly changing 21st-century environments and addressing the "needs of the globalised 21st century" (New curriculum in Zimbabwe, Pindula, September 26, 2016). Like Kenya and South Africa, Zimbabwe, a largely agro-based country in southern Africa, has also embarked on curriculum change.
To address this challenge, Zimbabwe introduced the ZCF 2015-2022 to shift from 'an examination centred education to an outcomes-based education that aims to equip learners with knowledge, skills, attitudes, and values to promote growth in the 21st century (MoPSE, 2016). Joynes et al., (2019) also argue that demand for 21st century skills in contexts such as Zimbabwe need to define these attributes in terms of life skills because they should respond to the challenges such as poverty, conflict, economic and environmental crisis.
The success of the ZCF (2015-2022) depends on the provision of electricity (a central invention in the second industrial revolution (2IR) in schools, which is not stable in most local schools, particularly in remote areas. Therefore, the absence of electricity prohibits the regular use of technologies such as computers and the internet that drive education in the 4IR. Joynes et al., (2019) propose 'learning progressions' as an essential strategy to measure the effectiveness of a curriculum. Additionally, they argue that if teachers are given incentives and opportunities to network, it will build morale and much-needed evidence that policymakers can use to improve the effectiveness of educational policies.
Furthermore, there has been much disgruntlement to the viability of the ZCF (2015-2002) from many sectors: parents, teachers, and the industry, political and academic circles who cite its lack of practicality. In addition, there is not enough evidence on how secondary schools in non-Western contexts including Zimbabwe are responding to the curriculum changes in the 21st century (Makonye, 2019). Additionally, little knowledge is also available on how Zimbabwean students understand and respond to the needs of 21st century education. For instance, several studies at the international level have looked at 21st century teaching and learning from teachers and educators' perspectives and learners separately (Kereluik et al., 2013;Tharumaraj et al., 2018;Mishra & Mehta, 2017;Brown, 2018). These studies have either looked at what teachers need to know about 21st century teaching and learning or what learners need to know.
To add on, these studies have focused mainly on the skills such as communication, creativity and innovation, collaboration and critical thinking. A knowledge gap exists since there seems to be very little evidence on students' views on all aspects of the 21st century as outlined by the P21 framework and 21st century teaching and learning in an African context (Joynes et al., 2019). Further, there seems to be a potential disconnection between the demands for 21st century skills at the international and regional levels. For instance, differences in socio-economic contexts and a lack of clear cut definitions that clearly outline approaches and evaluation of the effectiveness of policies aligned to the teaching of 21st century skills in LMICs are the primary causes.
The study, therefore, seeks to find out the views of the students (Form 4) of their 21st century teaching and learning from a sample of four secondary schools in Zimbabwe. The following research question reflects the preceding discussion: How are students in a sample of secondary schools in Zimbabwe responding to 21st -century imperatives shaping social, political and economic priorities in a changing environment?
The sub-research questions below will guide the study: • What are the Zimbabwe students' conceptions of their learning of 21st century skills? • Is there an association between race, gender, age, school type on the learning of 21st century skills? • What are the implications of the students' conceptions of 21st century skills learning for teacher education and curriculum policy in Zimbabwe?

Theoretical Background
Many debates which are Western oriented have been put on the table regarding educational reform in the 21st century. OECD (2019, p.11) state that a "non-linear, dynamic model" shift in curriculum design and learning is called for. For this shift, several reasons are given. First, the "static, linear learning progression model" is traditional and responds poorly to interdependent, dynamic and continuous systems that education has become in the 21st century. Second, the curriculum should recognise differences in students' learning paths and their unique talents in times of change. Third, the curriculum should recognise that students are individuals whose knowledge, skills, values and attitudes are unique and learn differently. Last, a shift from standardised assessment to a variety of assessments for various purposes should be adopted by schools to cater for multiple intelligences and not just one. These views echo the complex learning approach raised by Jess et al., (2011, p.3) that defies the traditional educational system, which "is based on linear and behaviourist learning models" and seemingly fixed knowledge. In addition, a complex system that is preoccupied with "change and adaptability in changing environments" is learner-centred and responsive. Of particular interest are Kereluik et al.'s (2013) meta-review of 15 reports, books, articles and an examination of strategies used by various frameworks to determine which knowledge is most important in 21st century teaching and learning and how teachers and educators view this knowledge. Their findings show that the frameworks and teachers agree that three forms of knowledge are vital in 21st century teaching and learning. These are represented in their 3 × 3 Model of 21st Century Learning framework, which outlines the foundational, meta, and humanistic types of knowledge as shown below. a) Foundational, which consists of the following constructs: Core content knowledge, digital/ICT literacy, cross-disciplinary knowledge. b) Meta, which consists of the following constructs: creativity and innovation, problem-solving and critical thinking, communication and collaboration. c) Humanistic, which consists of the following constructs: life/job skills, ethical/ emotional awareness, cultural competence. Kereluik et al., (2013) also concluded that while the 21st century frameworks represent emerging forms of knowledge, educational goals have not changed much in the 21st century. They add that education needs to consider the impact of technology in foundational, humanistic and meta-knowledge. Additionally, they state that "though the manner in which we represent knowledge and act upon it may change, the core idea of what we do as educators have not" (Kereluik et al., 2013, p.133). However, later studies by Mishra & Mehta (2017) and Tharumaraj et al., (2018) diverge on Kereluik et al.'s (2013) model, which balances the three types of knowledge for several reasons. In particular, Tharumaraj et al., (2018) queried that Kereluik's study did not consider learner perspectives about which skills and knowledge are to be considered relevant to the world of life and world in the 21st century when they formed their 3 × 3 model. Tharumaraj et al.'s (2018) empirical study showed that students placed more value on meta and humanistic skills and knowledge more than they did on foun-dational knowledge. For example, skills and knowledge such as problem-solving, critical thinking, communication and collaboration, life skills, leadership, technology/digital literacy, creativity and innovation were rated higher than other skills. In their concept note, the OECD (2019, p.102) defines values as "guiding principles that underpin what people believe to be important when making decisions in all areas of private and public life. They determine what people will prioritise in making a judgement and what they will strive for in seeking improvement". Attitudes are anchored on a person's values and beliefs, which impact his or her behaviour, giving birth to positive or negative attitudes determined by contexts and situations.
The OECD (2019) remarks that classroom and school environments of the 21st century are vital if they shift from being top-down to student agency and teacherstudent co-agency. This shift creates dynamic environments where students have different pathways in learning because they come to school already with different knowledge, skills, values and attitudes. Therefore, different assessments should cater to these differentiation elements, perhaps through a bottom-up process. Additionally, findings from the PISA 2018 assessment show that academic success is closely associated with support from parents, conducive learning environments and positive growth mindsets (Schleicher, 2019). In the words of Schleicher (2019, p.10), "history shows that countries with the determination to build a first-class education system can achieve this even in adverse economic circumstances and their schools today will be their economy and society tomorrow." The P21 and OECD (2019) point to learning environments as more than learning materials and maintenance of school grounds and buildings.
Some studies have also been conducted that assess the acquisition of the 21st century by gender. Snir et al., (2017) posited that the motivational and psychological concept of 'locus of control' could explain why females do better than males in these areas since females' 'locus of control' is said to be external to a larger extent compared to that of males. Accordingly, females have an external locus of control and their personalities are influenced more by the surrounding events than an internal locus of control, which believes that one's behaviour and talents influence the outcome of events.

Theoretical Framework
The two theories that framed the study are the powerful knowledge and the complexity theory.

Powerful Theory
The first lens is the concept of Powerful knowledge, a socio-epistemic concept developed by Young & Muller (2013) to describe education as promoting social equality (Elksne, 2020). Two overarching statements frame powerful knowledge: "that there is 'better knowledge' in every field, and (ii) that at the root of all decisions about knowledge in the curriculum is the idea of differentiation; that there are different types of knowledge" (Young & Muller, 2013, p230). Young & Muller (2013) distinguish between two types of knowledge: school/curriculum knowledge and everyday knowledge/experience that every child brings to school. The two types of knowledge have structural differences as well as different purposes. While everyday knowledge is context-dependent, school/subject knowledge is not. Therefore, when a child enters school, they bring their everyday knowledge, and the school curriculum helps them move beyond their everyday experience. To this end, teachers need to distinguish between curriculum and pedagogy clearly.
The school curriculum acts as a resource that teachers use to map their goals in the lesson, classroom, school and nation (Young & Muller, 2013). On the other hand, pedagogy is the 'how' or methods/means/strategies used by teachers through reconceptualisation to align every pupil's everyday experience as scaffolding for new concepts from the curriculum.
Such an experience on learners will enable them to view their everyday experiences in authentic ways (Young & Muller, 2013).

Complexity Theory
The second lens is the complexity theory which describes the ever-changing interaction between systems and their environments resulting in reorganisation of the system at every period of disturbance. In education, the elements of the system include diverse knowledge, cultures and contexts, students, individual sense-making, teacher-learner relationships, classroom dynamics (Davis & Sumara, 2009;Mason, 2008). In this technologically fast-changing, world reorganisation becomes an integral part of the complexity theory. Consequently, complexity applies to rapid change and curriculum change times (Amimo et al., 2014;Morrison 2008).
The critical components of a complex system are self-organising system, emergence, unpredictability/non-linear change, networks, continuous adaptation, sensitivity to initial conditions, connectivity and relations (Wood & Butt, 2014; Davis & Sumara, 2009;Mason 2008).
Self-organisation is the process of adapting to the new order (Forsman, 2015), and it rides on the edge of chaos, causing unpredictability and uncertainty. According to Morrison (2008), connectedness is an essential pillar of the complexity theory. Consequently, collaboration and communication are essential elements of the complex theory. In addition, Morrison (2008, pp.18-19) contends that as one moves towards the edge of chaos, " the more creative, open-ended, imaginative, diverse, and rich are the behaviours, ideas and practices of individuals and systems, and the greater is the connectivity, networking and information sharing (content and rate of flow) between participants. The natural consequence of this view of learning is an emphasis on the conditions to promote emergence, including motivation, enjoyment, and passion, cooperative and collaborative activity".
The enabling conditions for the complexity theory to take root are: internal diversity, decentralised control (bottom-up approaches), enabling constraints, neighbour interactions (communication and collaborations) and internal redundancy (shared characteristics). The complexity theory defies the traditional behaviourist approach to teaching, which promotes teacher-directed learning, but the complexity creates opportunities for student-directed learning, creating opportunities for collaboration, creativity, communication, and critical thinking (see Jess et al., 2011, Morrison, 2008.

The Elements of the Conceptual Framework and Their Definitions
The elements used in the survey instruments were adapted from the P21 framework. The survey has six themes for the student survey.
The conceptual framework comprises six skill sets which are: Learning and Innovation Skills (Critical Thinking, Collaboration, Communication, Creativity, and Innovation); interdisciplinary or cross-cutting issues (global connections; "financial, economic, business and entrepreneurial skills, civic literacy, health literacy and environmental literacy); Life and Career Skills (flexibility and adaptability, initiative and self-direction, social and cultural skills, productivity and accountability and leadership and responsibility); Information Media and Technology skills (Information and Media Literacy (IML) and Information Communication Technologies (ICT); 21st century standards and 21st -century learning environments.

Research Design
The study used a quantitative non-experimental correlation design.

The Instrument
The questionnaire was developed mainly from the P21 framework. In addition, some of the items for this instrument were adapted from "a teacher survey for measuring 21st -century teaching and learning; West Virginia 21st century teaching and learning survey" [WVDE-CIS-28] (Ravitz, 2014). The researcher then modified the items to suit the Zimbabwean context.
One reason for selecting the WVDE-CIS-28 was that in terms of reliability, it had proved to be excellent and reliable in other studies (for instance, std. alpha > 0.90, inter-item correlations > 0.58) (Ravitz, 2014).
The elements of the conceptual framework were divided into six skill sets. These were named Skill sets A to F. The individual components of the skill sets are shown below.
Skill set A -Learning and Innovation skills: Critical Thinking, Collaboration, Communication, Creativity, and Innovation. They are commonly called 4Cs, and they distinguish between learners ready to participate in the 21st -century environments and those who are not.
Skill set B -21st -century themes which refer to interdisciplinary or cross-cutting issues, which are: global connections; "financial, economic, business and entrepreneurial skills, civic literacy, health literacy and environmental literacy.
Skill set C -Life and Career Skills occur when learners select learning areas that interest them and assess their progress and learning from formative feedback (Hixson et al., 2012). The skills are: flexibility and adaptability, initiative and self-direc-tion, social and cultural skills, productivity and accountability and leadership and responsibility.
Skill set D -Information Media and Technology skills -using technology as a tool) which enable learners to function as knowledge consumers if successfully acquired and enable learners to contribute and impact policy decisions relating to using emerging technologies in societies. The two key terms are Information and Media Literacy (IML) and Information Communication Technologies (ICT), Skill set E -21st century standards involve embedding 21st -century skills in core subjects, resulting in deep learning with a consistently high-level testing system and constructive formative and summative assessments.
Skill set F -21st -century learning environments Learning Environments are the systems, mechanisms, and communities that inspire pupils and teachers to gain knowledge and skills the 21st century considers relevant (P21, 2009).
The questionnaire comprised 82 items based on six themes and 17 constructs. The constructs used in this study were Critical Thinking totals (CTtot), Collaboration totals (COLtot), Communication totals (COMtot), Creativity and Innovation totals (CItot) which fell under Skill set A; Global Connection totals (GCtot), Financial, Economic, Business and Entrepreneurial Literacy totals (FEBELtot), Civic Literacy totals (CLtot), Health Literacy totals (HLtot) and Environmental Literacy totals (ELtot) which fell under skill set B; Flexibility and Adaptability (FAtot), Initiative and Self-Direction skills totals (ISDtot), Social and Cross-Cultural Skills totals (SCCStot), Productivity and Accountability totals (PAtot) and Leadership and Responsibility totals (LRtot) which fell under skill set C; Information, Media and Technological Skills totals (IMTStot) which fell under skill set D; 21st century Standards (Stot) which fell under skill set E and Learning Environments (LEtot) which fell under skill set F.
The first section of the questionnaire contained biographical information on gender, race, age, type of school, and core subjects. The second section of the survey instrument contains six themes and 17 constructs, and 82 items to examine the 21st century teaching and learning in Zimbabwe. In the second section of the survey, respondents gave responses to every factor on a Likert-5-point frequency of use, starting with 1 (never), 2 (rarely), 3 (sometimes), 4 (often), and 5 (always).

Instrument Validity
Content, construct, and criterion validity are essential when designing a measuring instrument. Content validity occurs when the results given by an instrument reflect its intended use. Construct validity is the extent to which the instrument measures all aspects (breadth and depth) of the concept in question. Criterion validity measures how closely a test relates to other tests (Heale & Twycross, 2015;Hinkin, 1998) suggested adequate sampling of the domain (in this case, 21st -century education in Zimbabwe) to avoid measurement error in the content validity. This study adopted the P21 conceptual framework since it has been widely tested (Brown, 2018, Mishra & Mehta, 2017; Tharumaraj et al., 2018;Ravitz, 2014;Kereluik et al., 2013).
Content validity was achieved by using the P21 conceptual framework since it has been widely tested (Brown, 2018, Mishra & Mehta, 2017Tharumaraj et al., 2018;Ravitz, 2014;Kereluik et al., 2013). In addition, two other experienced researchers had to check whether the adapted questionnaire was fitting the Zimbabwean context. Furthermore, criterion validity was ensured by adapting some of the constructs in the questionnaire from Ravitz (2014) for skill set A and skill set E. Once adapted, the questionnaire items were carefully rephrased to fit the context of 21st -century education in Zimbabwe.
Construct validity measured by calculating convergent and discriminant validity of the constructs.

Methodology
First, the questionnaire was piloted to the Lower sixth form at one of the schools. Exploratory factor analysis was used to validate the questionnaire. Items that had poor loadings were discarded. The questionnaire was then administered to form four students in the four schools. One hundred nineteen males responded to the survey representing 50.4% of the sample. The remainder (n = 117) were females. 94.5% of the sample classified themselves as Black (n = 223), while 2.5% indicated they were either White (n = 6) or Coloured (n = 6). One responded indicated the 'Other" category. The majority of students were between 16 and 17 years old (n = 210) with 6.4% (n = 15) under 16 and 4,7b% (n = 11) 18 and over. There were three examination types that the students sat for in 2020. For Cambridge, there were 55 students, 6 for Cambridge "O" level and 175 for ZIMSEC "O" level. The validation process using exploratory factor analysis was repeated.
Four schools took part in the study. These were Pvt 1 with 61 students, Govt 1 with 60 students, Govt 3 with 55 students and Govt 2 with 60 students. The response rate was 100%.

Unidimensionality
Exploratory factor analysis (EFA) using the principal factor axis (PFA) was then carried out on each variable/construct to check for a single factor solution, signifying that the items in a construct probably measured the same thing (Unidimensionality). This was achieved by deleting items which did not contribute to a one-factor solution.

Sampling Adequacy
The sampling adequacy was calculated to test whether the sampling sample was adequate for factor analysis to take place.
The overall Kaiser-Meyer-Olkin (KMO) value greater was 0.891 and this value is greater than the minimum 0.5 thus indicating good sampling adequacy (Field, 2018).
The KMO values for each construct were greater than 0.5, indicating good sampling adequacy. The resultant factor loadings are shown in the Table 1 below. From Table 1 above, it is evident that the factor loadings for some of the items, Critical Thinking (CT1) and Communication skills (COM1), are poor and less than 0.5 (see Field 2018). All the other loadings loaded at values greater than 0.5 indicating high convergent validity of the items (Field, 2018).

Reliability
• Cronbach's alpha, Composite Reliability (CR), Average variance extracted (AVE) and Corrected item-total correlation were used to measure reliability.
The Cronbach's alpha measures the internal consistency of each scale. This was calculated for each construct. Composite reliability (CR) and Average variance extracted (AVE) measured construct validity (to verify whether the items measured the intended construct) (Ellis, 2017). Construct validity consists of convergent validity (correlated items in a construct) and discriminant validity (items in different constructs are poorly correlated).
Corrected item-total correlation measures convergent and discriminant validity (CR and AVE are elements of construct validity). Table 2 below shows the CR, AVE, corrected item-total correlation squared and Cronbach's alpha values for each construct.
In this study, all constructs had CR values greater than 0.7, indicating good reliability except for Critical Thinking and Communication skills, which had CR values less than 0.7 (Hair et al., 1998). Although some AVEs values were smaller than 0.5, the corresponding CRs values exceeded 0.6. Thus indicating convergent validity 0.50 (Fornell & Larcker, 1981).
For discriminant validity, all the AVEs values were greater than their matching corrected item-total correlation coefficients squared, indicating good discriminant validity.

Factor Rotation for the Constructs
A factor analysis extraction (principal factor analysis with oblique Promax rotation) of 17 constructs revealed a three-factor solution which accounted for 61.569% of the total variance. The scree plot confirmed that three factors were retained after extraction. Table 4 below shows the loadings of the 17 variables on the three factors extracted with factor loadings suppressed at 0.3 thus affecting the Communication totals (COMtot) construct, which had loadings less than 0.3. (see Table 3 below in the pattern matrix. The higher the value of the loadings, the greater the reliability of the factor (Field, 2018).
All the factor loadings for the variables clustered somehow randomly under three factors, as shown in the diagram above. This result suggests that three factors underline the conception of 21st century skills in Zimbabwe for students.

Table 1
The factor loadings for each construct scale

Confirmatory Factor Analysis
Confirmatory Factor Analysis (CFA) verified the fit of the measurement model during exploratory factor analysis. AMOS 26.0 was used for the analysis.
The latent variables for the first model before the modification are shown in Fig. 1 below.
• The first confirmatory model fit was reasonable before any modifications were applied.
The indices used to assess the model fit were chi-square to degrees of freedom (CMIN/DF), the Comparative Fit Index (CFI), Tucker Lewis Index (TLI), incremental fit index (IFI), Standardised Root Mean Residual (SRMR), and Root mean square error of approximation (RMSEA). The fit indices were CFI -,928, TLI -,915, IFI -,929. Although these values were acceptable, they were less than the threshold of 0.95,. SRMR was 0535 and RMSEA was, 078 and both values were less than the threshold of 0.08.
The model's fit was then improved through modification by co-varying items of the factors. Figure 2 below shows the latent variables of the final model after modification.
F1, F2 and F3, were given names depending on their relevance to skills and values in the P21 framework. F1 was named learning skills and values since the constructs cover health literacy, environmental literacy, global connections and business literacy. F2 was named learning processes since the constructs cover learning environment, collaboration, and media aspects. F3 was named self-management since the constructs cover productivity and accountability, leadership and responsibility and social and cross-cultural skills.
After applying model evaluation and modification, the standardised estimates improved to almost greater than 0.7 for most variables. According to Field (2013), loadings greater than 0.7 are excellent. The standardised regression weights of the default model are shown in Table 4 below.

Reliability
Reliability was measured by considering CR, AVE and the square root of the AVE (see the bold numbers presented diagonally in the table below). Table 5 below shows the convergent and discriminant validity of the constructs used.
The CR values are greater than 0.7, and the AVEs are greater than 0.5 signifying convergent validity (Hair et al., 2006). The square root of the AVE (bolded diagonally) for F1 is greater than their inter-construct correlation for F2 and F3, but for F2 and F3, the inter-construct correlation is smaller than their inter-construct correlation, which signified some poor form of discriminant validity. According to Fornell & Larcker (1981), discriminant validity is achieved when the square root of the AVE is greater than their inter-construct correlation.

The Model fit Measures
The resultant final model had excellent fit indices.
The CFI was equal to 0,950, which is greater than 0.90; the RMSEA was equal to 0,067, which is less than 0.08, indicating a good model fit to the data. The TLI was   Fig. 2 The confirmatory factor analysis model equal to 0.938 a value greater than 0.90, and the SRMR was equal to 0,0483, which is less than 0.08, also indicating a good model fit to the data. Generally, the analysed model had a good fit as indicated by the fit indices, and thus, the measurement models were confirmed (see Hu & Bentler 1999).

The Second-Order Confirmatory Factor Analysis
The poor discriminant validity above suggested that second-order factor analysis was necessary to explain the latent variables (F1, F2 and F3). The second-order factor was named Conception of 21st century skills. The results of the second-order analysis appear in Fig. 3 below.
The values in Table 6 below further confirm the validity and reliability of the Conception of 21st Century skills (21st_Century_S). LSV denotes learning skills and values, LP denotes learning processes, and SM denotes self-management.
Table 7 below shows that the Conception of 21st century skills (main construct) on all subcontracts is significant.    The p-values in Table 7 above show that all components in the Conception of 21st Century skills model are highly significant since their respective p-value is lower than 0.01.
All the model fit indices achieved the required level. The results show that the students' conception of 21st century skills loads very well on its three sub-constructs. The factor loadings of learning skills and values, learning processes and self-management are 0,963, 0,864 and 0,813, respectively. Further, the subcontracts also loaded highly. Consequently, the suggestion that the Conception of 21st century skills for students in Zimbabwe comprises three elements: learning skills and values, learning processes, and self-management is well supported.

Results
The three emerging factors from the data analysis were learning skills and values, self-management and learning processes. The associations between independent and dependent variables (learning skills and values, self-management and learning processes) were investigated using a one-way between subjects' Analysis of Variance (ANOVA) test provided the Levene's test for the homogeneity of the variances was equal and alternatively using the robust tests of equality (Welch, Brown-Forsythe) for non-equality of homogeneity.

Learning Processes
The elements of Learning Processes are Information, media and technological skills (IMTStot), 21st century standards (Stot), 21st Century Learning Environments (LEtot) and Collaboration (COLtot) skills.

Self-Management
The elements of Self-Management are Productivity and Accountability skills (PAtot), Leadership and Responsibility (LRtot), Initiative and Self-direction skills (ISDtot) and Social and Cross-cultural (SCCStot) skills.

Summary of Findings
In this section, the results that had a statistically significant association with selfmanagement, learning processes and learning skills and values are summarised.
• The significant association between independent variables and dependent variables. Table 8 below summarises the associations between the independent variables (gender, population group, age, type of school and exam type) and dependent variables (learning skills and values, learning processes and self-management).
• Comparing zcores for self-management, learning skills and values and leaning processes.
From Fig. 4 below, self-management is the most significant 21st century skill, followed by Learning skills and values and then learning processes for the students in Zimbabwe as represented by the four schools that took part in the study.

Discussion
This research had three aims. First, it sought to find out the Zimbabwean students' conceptions of their learning of 21st century skills. Second, the study aimed to examine if there is an association between race, gender, age and school type on the learning of 21st century skills. Third, the study aimed to find out the implications of the students' conceptions of 21st century skills learning for teacher education and curriculum policy in Zimbabwe. The Students in this study perceived self-management followed by learning skills and values and, thirdly learning processes as necessary.
The varying levels of endorsement make sense in that students endorse what they want to learn before they endorse how they learn. The grouping of 21st -century skills into three categories is consistent with findings from Kereluik et al., (2013). However, these elements of learning skills and values, learning processes, and selfmanagement do not align with Kereluik et al.'s (2013) humanistic, foundational, and meta-knowledge.
In their meta-analysis study, Kereluik et al., (2013) posited that the 21st century skills domains/themes were divided into three categories: foundational (core content knowledge, digital/ICT literacy, cross-disciplinary knowledge), meta (creativity and innovation, problem-solving and critical thinking, communication and collaboration) and humanistic, (life/job skills, ethical/emotional awareness, cultural competence). In this study, the respondents indicated that the categories of knowledge do not have equal significance in the knowledge needed for the 21st century. Findings from this study, however, are partly consistent with those of Tharumaraj et al., (2018), in Malaysia, on learners' perspectives of 21st century skills in that students do not view the elements/themes of the 21st century skills to be equally important because some elements are seen to be more critical than others. However, these findings differ from Tharumaraj et al.'s (2018) aspects in ranking these elements as shown in Table 9, where problem-solving/critical thinking was rated the most important, followed by communication and collaboration and life/job skills/leadership. Also shown in Table 9, the current study showed that critical thinking is in the second category while collaboration is in the third category for the students. The order of the findings differs from the empirical study by Tharumaraj et al., (2018). The probable reason is more likely to be that different instruments have been used and quite different populations (mature students versus school students. Several points could be made. Firstly, there are differences between younger and older students used in the two studies. Secondly, there are differences in contexts (Zimbabwe versus Malaysia), and thirdly, different instruments produce different results (the current study used the P21 while Tharumaraj et al., (2018) did not. Yet, overall, the point is that different versions of 21st century skills are viewed as necessary, although viewed from different perspectives and for different reasons.

The Students' Conceptions of Their Learning 21st Century Skills
The first aim of this study pertained to the students' views on their learning of 21st century skills. The students' results for the learning skills and values, learning processes and self-management are discussed below.

Self-management
Self-management comprised the following elements: Productivity and Accountability skills (PA), Leadership and Responsibility (LR), Initiative and Self-direction skills (ISD) and Social and Cross-cultural skills (SCCS). It was the most strongly endorsed factor in the model. All the elements in this group are from Skill set C (Career and Life skills) of the P21, and students consider them to be of most worth in the Zimbabwean context. The students' results showed a high value on their career and life skills which differed from Ahonen and Kinnunen's (2015) study of Finnish students who ranked social skills and collaboration highest. In Tharumaraj et al.'s (2018), the study of mature students' life/job skills occupied third place (middleorder). It is important to note that the Zimbabwe economy has been unstable for decades, and this could be one reason learners are preoccupied with civic, productivity, adaptability, self-direction, social and cross-cultural issues. Their future, with or without an education, is unstable. There is a plausible explanation in Joynes et al., (2019), who discuss global and regional differences in the demand for 21st century skills. Their discussions highlight the differences within regional contexts/developing countries and underdeveloped countries because of priorities in policy and labour market capabilities. The same goes for future demand as well. For example, LMICs like Zimbabwe are primarily informal, agro-based and require little high-tech skills. In this study, the students' choice of self-management is consistent with the literature of other LMICs (South Africa and Kenya) (Care et al., 2017 as cited in Joynes et al., 2019) where attributes linked with life skills, in particular, community, society and citizenship -are noteworthy constructs globally and are highly considered by stakeholders at the state level. The students in Zimbabwe find life skills such as productivity and accountability, initiative and self-direction, leadership and responsibility and social and cross-cultural skills to be more important in their context where demand for these skills is higher than attributes of globalised definitions of 21st century skills such as communication, critical thinking, collaboration, creativity and innovation and technology and computer skills (Joynes et al., 2019).

Learning Skills and Values
Learning skills and values comprise the following elements: Civic Literacy (CL), Flexibility and Adaptability (FA), Environmental Literacy (EL), Health Literacy (HL), Global Connectivity (GC), Financial, Economic, Business and Entrepreneurial Literacy (FEBEL), Creativity and Innovation (CI) and Critical Thinking (CT) skills. CL and FA are elements of Skill set C in the P21, while EL, HL, GC and FEBEL are cross-disciplinary themes and fall under Skill set B of the P21. Finally, CI and CT are learning and innovation skills in Theme A. In this study, students have grouped three skill sets from the P21 (Skill sets B, A and C) into a new skill set named Learning skills and values, thus, combining Skill sets B, A and C into learning skills and values. As pointed out by OECD (2019), skills and values will prepare students for emerging occupations. The students' endorsement suggests that they are also aware of essential skills and values that will support them. It is worth noting that issues dealing with health, the environment, financial, economic and entrepreneurial literacy in Zimbabwe are central considering the prevailing deteriorating economic, political and social conditions. In this country, the learners appear to have become alert about these issues because of the ailing economy, the successive droughts and the non-existent health sector. Additionally, learners have learnt entrepreneurial skills to source funds for their education or cater to their social needs. According to Joynes et al., (2019), developmental settings such as LMICs, of which Zimbabwe is one, define life skills as attributes that ensure resilience in the face of poverty, conflict and economic and environmental crisis. This definition may explain why students are aware of these learning skills and values because such issues define their context.

Learning Processes
The elements of learning processes are information, media and technological skills (IMTS), 21st century standards (S), 21st Century Learning Environments (LE) and Collaboration (COL) skills. In the P21, IMTS is Skill set D; S is Skill set E; LE is Skill set F, and COL is part of learning and innovation skills (Skill set A). According to the student's views, these constructs are ranked third after self-management and learning processes and values.
The elements in this skill set reflect the gaps in accessibility in most Zimbabwean schools. For instance, the poor technological infrastructure in most Zimbabwean schools could explain why digital skills are ranked lowest in this study and higher by Malaysian pre-university students (where Tharumaraj et al.'s (2018) took place, which has a better technological infrastructure (Tharumaraj et al., 2018) as compared to Zimbabwe. This dramatically limits the students to evaluate something highly that they do not have access to. Similarly, learning environments and 21st century standards fall into this skill set which the students ranked last. According to the OECD (2019), classroom and school environments of the 21st century are vital if they shift from top-down to student agency and teacher-student co-agency. The P21 and OECD (2019) point to learning environments as more than learning materials and school grounds and buildings maintenance. Complexity theory also stresses the impor-tance of sensitivity of initial conditions in complex systems because the conditions determine the entire learning process (Morrison, 2008). The 21st century standards involve embedding 21st century skills in core subjects, resulting in deep learning with a consistently high-level testing system and constructive formative and summative assessments (P21, 2009). Collaboration was considered the least essential element since the students ranked learning processes third. This finding validates the lack of "neighbourly interactions" such as brainstorming together and questioning strategies employed to generate collectivist ideas or knowledge in decentralised-controlled classrooms that create conducive environments that foster collaboration (Wood & Butt, 2014). Hence, teamwork and systematic thinking among the team members can educate learners about themselves and their place in society, resulting in them changing. This collective engagement raises learners' self-awareness, which leads to an appreciation of each other. The 4Cs of which collaboration is a part is the foundation from which the reconceptualisation process occurs. If the students have endorsed it last, it suggests that they are probably unexposed to collaboration strategies.

Communication
Unlike other studies, communication was not endorsed strongly enough to be included in any endorsed factors. Communication skills loaded very poorly (factor loadings of less than 0.3) and were discarded. According to Field (2018), such poor factor loadings must be eliminated as statistically insignificant. The use of teachercentred strategies and lack of straightforward student approaches could mean that no collaboration between teachers and students occurred. Therefore, communication skills need fostering since they encourage complex social skills such as persuasion and negotiation, which are essential in caring professions (OECD, 2019).

Association between race, gender, age, school type and examination type on the learning of 21st century skills for the students
The second aim pertained to the possible association between race, gender, age, school type and examination type with the learning of 21st century skills. The associations are discussed below.

The association between gender and self-management, learning processes and learning skills and values for students
Gender had a statistically significant association with self-management and learning skills and values where the mean scores for the females were greater than those of the boys. The elements of learning skills and values are civic literacy, flexibility and adaptability, environmental literacy, health literacy, global connectivity, financial, economic, business and entrepreneurial literacy, creativity and innovation and critical thinking skills. The elements of self-management are initiative and self-direction, productivity and accountability, leadership and responsibility and social and crosscultural skills. Thus, the female respondents perceived learning skills, values and self-management as occurring significantly more often in their core subjects than in male participants. Most female respondents in this study come from the three government schools (Govt 3, Govt 1 and Govt 2). Gender had no statistically significant association with learning processes. Snir et al., (2017) posited that the motivational and psychological concept of 'locus of control' could explain why females do better than males in these areas since females' 'locus of control' is said to be external to a more considerable extent compared to that of males. Accordingly, females have an external locus of control and their personalities are influenced more by the surrounding events than an internal locus of control, which believes that one's behaviour and talents influence the outcome of events. However, for learning processes, gender had no effect.

The association between exam type and self-management, learning processes and learning values and skills for students
Examination type had a statistically significant association with learning skills and values, learning processes and self-management. The students sitting the Zimbabwe Secondary "O" levels (ZIMSEC) examinations have a significantly higher mean score than participants involved with the International Cambridge Curriculum (IGCSE).

The association between school type and self-management, learning processes and learning values and skills for students
School type had a statistically significant association with learning skills and values, learning processes and self-management. First, the association between school type and learning skills and value is discussed. Learning skills and values consist of the following elements: Civic Literacy, Flexibility and Adaptability, Environmental Literacy, Health Literacy, Global Connectivity, Financial Entrepreneurial Business and Economic Literacy, Creativity and Innovation and Critical Thinking.
For learning skills and values, the mean scores for Govt 3 were greater and statistically significant than that of the other three schools (Pvt 1, Govt 1 and Govt 2). Govt 3 had a statistically higher mean score over the other schools because it has created opportunities for emergence (a product of self-organisation in the complex theory). According to Horn (2008) and Mason (2008), this emerging self -organisation is the primary driver of the complexity theory and is characterised by "adaptability, open systems, learning, feedback, communication and emergence". Self-organisation allows the system to renew itself with time or self-organise whenever the system has been disturbed or put on the edge of chaos (Morrison, 2008). The mean of Govt 2 and Govt 1 was greater and statistically significant than that of Pvt 1. Pvt 1 is a private school and has an economically advantaged background to the other schools. Surprisingly, it is lagging behind these economically disadvantaged schools. Perhaps, adverse economic circumstances are critical drivers to success in government schools. These economic hardships ensure that students value skills such as civic literacy.
However, there was no statistically significant difference between Govt 1 and Govt 2. The conditions of the two schools are similar in that they are both government schools following the same syllabus.
Second, the association between school type and learning processes is discussed. The elements of the Learning Processes are Information, Media and Technological Skills, 21st century Standards, Learning Environments and Collaboration. For learning processes, the mean score of Govt 3 was statistically significant and greater than those of Pvt 1 and Govt 1 but not statistically significant to the score of Govt 2. Govt 2 and Govt 3 are schools in the same high-density area of Chinhoyi town, which could explain their similarity. However, as previously mentioned, Pvt 1 has better resources, and one would have expected it to be better than Govt 3, but it is not. According to the OECD (2015), learning environments are places "where the kids know each other and know their instructors, not just academically but as people." Among the benefits of such an environment are character traits such as friendliness, openness, cooperation, interaction and reduction of tensions, inattentiveness and bullying, to mention a few. While it was evident that Govt 3 faces challenges with information, media and technological skills, it is worth noting that information, media and technology is not the only or significant enabler for conducive and collaborative learning environments.
The scores of Govt 1 were lower and statistically significant compared to those of Govt 2. The learning environment for Govt 2 could be better than that of Govt 1. There were no statistically significant differences between Govt 1 and Pvt 1. Perhaps the learning processes highlighted here go beyond independent variables such as high status and resources. From Young and Muller's (2013) powerful knowledge perspective, one could say Govt 2 is doing a better job using the learners' everyday knowledge as scaffolding for learning new knowledge in their core subjects.
Pvt 1 has a better technological infrastructure that is not used as a good learning environment enabler but a stand-alone subject in terms of information, media and technology. If the schools in Zimbabwe want to move their learners into the 21st century, they have to reimagine new ways of learning, working and responding to global issues involving their learners' views.
Third, the association between school type and self-management is discussed. The elements of self-management are productivity and accountability, leadership and responsibility, initiative and self-direction and socio and cross-cultural skills.
In self-management, the scores of Govt 3 were greater and statistically significant than those of Pvt 1, Govt 1 and Govt 2. It follows that when one acquires learning values and processes that are superior and effective, one can manage individual learning. This is the case for Govt 3 where the mean scores have been statistically significant and higher than those of the other schools in self-management. The students can now self-manage.
Overall, Mason (2008) posited that sustained effort and interventions are needed to trigger emergence growth. Student-centred approaches, bottom-up processes and decentralised control are essential in creating dialogue, creativity and connectedness channels for greater educational outcomes. However, it is worth noting that there is still a long way for all these schools to go before these elements of the complex theory are pervasive in their learning. Emerging results also show an interesting pattern where schools from disadvantaged backgrounds have a superior understanding of learning skills and values, processes and self-management skills to those with superior infrastructure and economic backgrounds. In the words of Schleicher (2019), "history shows that countries with the determination to build a first-class education system can achieve this even in adverse economic circumstances and their schools today will be their economy and society tomorrow". The same goes for the schools in this study as well.

The Association Between age and Learning Values and Skills
Age had a statistically significant relationship with students' views of learning skills and values.
The results show that older students think learning skills and values are more important than younger students. One reason could be that as one grows older, one becomes mature and has a better understanding of skills and values related to one's health, environment, civic duties and finances.

The Association Between race and self-management
Race had a statistically significant association with self-management. In all three themes, the mean for blacks was greater than the mean for 'Others' in self-management. Most of the black population used in this study comes from adverse socioeconomic backgrounds, and as such, low-income backgrounds push learners to do better and foster a growth mindset to aspire to improve oneself through education (Schleicher, 2019). Additionally, studies carried out by Care et al. (2017 as cited in Joynes et al., 2019) articulate the same notion that students in similar contexts as Zimbabwe (South Africa and Kenya) value attributes that promote one's well-being/ self-management, such as confidence, independence and responsibility, productivity, ethics and happiness.

The Association Between Population Group and Learning Skills and Values and Learning Processes
Population groups had a statistically significant association with learning skills and values, with the mean for blacks being greater than the mean for 'Others'. As mentioned earlier, the majority of the black population came from disadvantaged backgrounds, making them much aware of the need for health, civic and environmental literacy compared to students from a better socio-economic status who may not view these aspects as primary needs.
Besides, different races have different value systems. Another reason could be because blacks have the zeal to prove that they are better or equal placed citizens. Furthermore, they could have higher aspirations and motivation to move from disadvantaged communities to better communities through education. In contrast, the white and other races may already have the privilege and do not take education as a gateway to success.
The population group had a statistically significant association with learning processes, with the mean for blacks being greater than for 'Others' in learning processes. Learning processes consisted of information, media and technological skills, 21st century standards, learning environments and collaboration.

The implications of students' views of 21st century skills learning for teacher education and curriculum policy
The third aim pertained to the implications of the students' conceptions of 21st century learning for teacher education and curriculum policy.
While literature is available to outline the need for educational reforms due to the rapidly changing environments caused by globalisation, the exponential growth in technology and the needs of 4IR, the different socio-economic conditions existing in different contexts mean that the definitions and skills needed in those contexts are also different. A one size fit for western and regional or local contexts will not suffice. Therefore, the 21st century skills learning curriculum and pedagogy must also be different. Besides, as Joynes et al., (2019) articulated, the impact of globalisation and advances in technology is heterogeneous. It follows then that the needs and demands of 21st century skills differ globally and regionally. It would be unfair to expect contexts barely managing to reach basic numeracy and literacy. STEM and STEAM need to tackle higher-order skills such as problem-solving, information media and technological skills, critical thinking, and collaboration as defined by frameworks at international standards. Instead, local contexts and policymakers should redefine these same skills to suit their social imaginaries. For Zimbabwe, 21st century skills should be aligned to the agro-based context and respond to the demands of an informal economy. Life skills that address the well-being of the students and the teachers should be enforced to equip learners with self-management skills, and the learning processes should be aligned to suit the attributes of self-management and learning skills and values.
Upskilling the learners for the demands of the future economy should be done within the context they live. Leapfrogging the learners to acquire 21st century skills to respond to the knowledge economy may be an absolute, imagined dream in the strictest sense for Zimbabwe.
Most countries engaging in educational reforms are doing so to respond to labour market needs, and Zimbabwe is no exception. Its vision and educational reform in the ZCF 2015-2022 are to shift focus to address the "needs of the globalised 21st century" (New Curriculum in Zimbabwe -Pindula, 2016). Before Zimbabwe broadens its curriculum to encompass the global needs, it must address the local needs and align reforms to reflect the views of the beneficiaries.
Educational policy in Zimbabwe will need to ensure that curriculum, pedagogy and learning assessments for the teaching of 21st century skills are aligned. This will aid in developing effective educational reform that aligns the teaching of 21st century skills that marry the nation's social, economic, and political climate. By so doing, powerful knowledge will be used to bring about social change.

Recommendations
The researcher recommends hybrid curricula framed by complexity and powerful knowledge theories. There is little reconceptualisation and transformative learning in the schools from the findings. There is a need to include student-centred approaches that will foster nurturing of the 4Cs that are currently lacking. Conditions giving birth to emergence [new behaviours] should be promoted in schools. These are selforganisation, collectivity, sensitivity to initial conditions, neighbourly interactions that will enhance the application of the 4Cs.
With more application of the 4Cs, the researcher envisages the development of more skill sets other than self-management, learning skills and values and the Learning processes. For instance, a 4Cs theme must evolve with hybrid curricula. At the moment, the students value self-management skills ahead of learning skills and values and learning processes. With the incorporation of hybrid curricula based on the reconceptualisation of content and transformative pedagogies, the researcher anticipates that the students should valuing the themes equally in future.
There exist gaps between males and females in command of 21st -century themes. Consequently, empowerment programmes currently catering for females must also include boys to level the playing field.

Limitations
The study was only limited to Form 4 students in four secondary schools in one province. The study also took place over a single point in time. A longitudinal study would be adequate to see the cycles of emergence and self-organisation. A large sample would have been desirable to improve the sampling adequacy.

Theoretical and Practical Contributions
The study extended the applicability of the P21 framework to suit the Zimbabwe context. In Zimbabwe, the "P21 framework" is currently represented by three skill sets: learning skills and values, learning processes, self-management.
The results are also significant primarily for high school head teachers and teachers in how best they can implement and plan for effective policies, which should pay attention to the cross-cutting issues that can boost the learners' understanding of 21st -century skills inside the classroom. High school head teachers must also recognise that the skills of the 21st century are impacted by variables such as gender, age, race, exam type and school type to create a more equitable educational system. This exposure is significant as a social justice measure because it highlights that educators need to ensure that all students receive the same level of opportunity and equity no matter one's race, school type, exam type, age or gender.
The findings of this study regarding student perceptions of the importance and practice of 21st -century skills are particularly pertinent as educational leaders continue to develop theories on effective school instruction and create professional development for teachers in the future.
The study's findings can provide educational leaders with an understanding of the educational experiences of Ordinary Level and IGCSE students and their perceptions of learning and teaching using the P21 Learning Framework. More specifically, the study revealed which skills/elements students think are most important and how improvements can be implemented based on the baseline findings. Last but not least, policymakers will be able to make informed decisions about implementing 21st century skills.

Conclusion
Students in Zimbabwe endorsed self-management skills above learning skills and values and learning processes. The findings have shown that while the Western world from which the P21 originated perceives knowledge and skills in the 21st century to be grouped into six skill sets (the Zimbabwe students perceive these skills as three different skill sets as shown in Table 9 above. It is worth noting that Zimbabwe's ZCF 2015-2022 in 2017 is very similar to the P21, a western cultural concept that may not apply in different African cultural contexts. Consequently, cultural differences and economic, political and social levels will cause Zimbabwe's different 21st century learning framework. In this study, the students ranked self-management first, indicating no differentiation between what the students bring to school (everyday experiences) and what the school curricula offer. In this study, self-management is everyday knowledge. The elements of self-management are productivity and accountability, leadership and responsibility, initiative and self-direction and socio and cross-cultural skills. The students agree with Young & Muller (2013) that learning should start from the known (self-management) to the unknown (specialised knowledge in specific disciplines, which seems to be non-existent or not crucial as per the students' conceptions).