Critics of inclusion continue to argue that the concept is a source of much confusion among scholars and other stakeholders alike. On one hand, other critics believe that inclusion, its goal and intervention models arise from the need to go beyond the physical integration of learners with SEN in regular education to the quality of educational outcomes (Hallahan, Kauffman & Pullen 2012; Warnock & Norwich 2010). On the other hand, scholars like Terzi (2014) is of the view that, for some learners, special education schools is one of the options to meet their needs. Cigman (2007) argues that children with SENs differ remarkably and ‘one size fits all’ principle should not be applied to them.

Some prominent detractors of inclusion such as Scruggs and Mastropieri (1996) and Kauffman et al. (2018) have continued to argue that most inclusive practices emphasise social integration to the detriment of the acquisition of content knowledge and skills when the goal of educational programming is not only socialisation but the attainment of meaningful outcomes for all learners. Similarly, Lindsay (2007:370) claims that inclusion should go beyond mainstream but should include meaningful participation in the education process. To this end, Lindsay (2007) proposes learners with special needs should be identified and provided for appropriately.

In addition, Florian and Spratt (2013:121) contend that using the inclusive pedagogy approach entails rejecting the labelling of learners based on ability. They also argue that diversity should be seen as a strength and not as a problem. Others suggest that the focus should be not only on the students with disabilities but on all students and then consider how to accommodate those with disabilities into mainstream education (Loreman et al. 2014; Naicker 2018).

Hornby (2011) presents the sources of the confusion regarding inclusion and refers to these as ‘inclusion confusions’ (see Figure 1).

What Figure 1 depicts is that, if the implementers of inclusive education are still confused as highlighted by Hornby (2011) then diversity will be seen as a problem and not a strength. Confused implementers of inclusive education would still focus on students with disabilities and how to reasonably accommodate them in inclusive education settings (Hornby 2011). Instead, focus should be on all students with and without disabilities. For inclusive education to be effective, according to Hornby (2011), all things should be considered: (1) achievable inclusive education goals (among them not leaving any child behind), (2) identification of needs and support needed based on research evidence, (3) resources, (4) being realistic about what is possible and not possible.

Curriculum reform to facilitate the inclusion of learners with SEN and other marginalised groups of learners generally accompanies many inclusive policy documents. In the Zambian case, item ‘vi’ (6) in ‘Educating our Future’ includes a call to address the need for curricula change to support learners with SEN (Republic of Zambia Ministry of Education 1996). However, how to develop pedagogy and curricula that will be inclusive and beneficial for all learners is easier said than done (Davis & Florian 2004:34). The need for curriculum reform is especially challenging because the focus on performance as a marker of success in mainstream education contexts persists. Wolfe and Hall (2003) have previously argued that inclusion of students with SEN necessarily requires redesigning of curriculum and the provision of appropriate classroom adaptations. These adaptations include, but not limited to: (1) seating arrangements, (2) rearrangement of the layout of the classroom, (3) limiting visual clutter on walls, (4) sensory stimulation with soft or noisy music, (5) reduction of noise in a classroom to accommodate learners with visual impairments, (6) controlling lighting to accommodate partially sighted learners and those with autism, (7) adapting furniture by lowering chairs or securing desks and creating slant boards throughout the classroom for writing support, and (8) adapting door handles for learners with orthopaedic impairments, cognitive and/or developmental delays.

A focus on improving the participation of learners with SEN is less appealing as they are considered to be a liability on high-stakes tests reports (Juneja 2018:21). Yuen et al. (2019) argue that a common curriculum following the principles of Universal Design for Learning (UDL) should allow for teachers to monitor the progress of all the learners. Hence, Flood and Banks (2021) suggest that UDL should not lose sight of the importance of monitoring learner outcomes. In this process, Mukminin et al. (2019) posit that this access to the curriculum by all learners should include those without SEN. However, it is worth noting, as Meier and Rossi (2020) argue, that UDL does not imply a ‘one size fits all’ curriculum document but a programme of learning which removes barriers to learning for all learners so that the curriculum becomes an integral part of the framework. In practice, Theoharis and Causton (2014) provide some practical steps and considerations to promote efficient and effective curriculum reform and implementation which are:

Curriculum reform has not been at the forefront in the quest for inclusive education. Although in the Zambian case, all curriculum documents after the first education reforms of 1997 have a specific section relating to equity and inclusion of learners with SEN albeit with no details about how to differentiate for exceptionality of learners with disabilities. In our view, education reform calls for a paradigm shift from separate standards, curricular and accountability systems between special education and mainstream education. It is therefore important, as Conn and Hutt (2020) posit, that all stakeholders are committed to the inclusion of all learners in the curriculum renewal and implementation. They argue that beneficence for all should be the goal through a curriculum that affords multiple pathways to learning. This includes the need for mainstream public schools to be conducive to SEN learners and that education must ensure equal access to training, skills development, equal opportunities as well as career paths of all learners.

Purposive sampling was used to select a small group from a larger group with similar characteristics considered to be knowledgeable of, and informative about the phenomenon of interest; in this instance the provision of inclusive education to learners with SEN in Zambia (Maxwell 2013; Onwuegbuzie & Leech 2009; Onwuegbuzie et al. 2010; Terrell 2012). The study participants came from seven (Lua pula, Lusaka, Muchinga, Western, Southern, Central & Copperbelt) provinces of Zambia. A purposively sampled small group also participated in the quantitative part of this study. The aim was to gather valuable information and to assess the extent of the education reform. The target population which comprised of teachers, administrators and Education Support Officers was deemed to meet the criteria related to teaching learners with SEN, different provinces, gender, age, post description, responsibilities and duties. As a result, the largest number of respondents were SEN teachers and administrators currently involved in providing special education in Zambia.

There were seven focus group interviews which were held at seven different sites. The intention of the researchers was to interview 38 participants but because of circumstances beyond their control, they ended up interviewing 18 participants only.

Zambia policy document on curriculum reform was downloaded from the Internet. In analysing document, special attention was given to units of meaning and set of categories. Researchers made notes from available documents. Researchers ensured authenticity by being open minded and referring to credible Zambian government documents. As part of document analysis, the following documents: Educating our Future (Republic of Zambia Ministry of Education 1996) and Zambia Education Curriculum Framework (Zambia MESVTEE 2013) were targeted.

Policy documents were analysed to determine whether issues of curriculum reform were taken into high esteem by the Government of Zambia. Furthermore, the researchers wanted to be sure whether those policies had clear inclusive education implementation plans.

LeCompte (2000:146) asserts that analysis required turning data into results. As a result, research results cannot be accurate or reliable if pieces of data are incomplete or biased. Kothari (2004:122) suggests that data, after collection, be processed and analysed in accordance with the outline laid down for the very purpose of the research plan. Analysis also implies editing, coding, classification and tabulation of collected data.

Interview notes and questionnaires produced a large quantity of data that needed to be summarised and interpreted. Common threads, themes and, patterns were identified. Thematic data analysis for qualitative data was employed. The following thematic data steps were considered:

In light of this idea, the researcher planned to be wholly immersed in the data so as to be familiar with the information or data and proceeded to systematically synthesise, organise, analyse, transcribe, segment and code data, eventually findings were then be presented. The researcher: (1) organised and prepared raw data for analysis, those were notes from archival documents and transcripts from interviews; (2) engaged or read data to get a general sense and reflected; (3) began by coding texts into consumable segments; (4) used codes to thematise or categorise for analysis; (5) presented how themes would help narrate a qualitative description; and lastly, (6) interpreted the collected data.

All 150 questionnaires for SEN teachers and 50 questionnaires for pre-service SEN teachers were received. Soon as data were collected, and while it was still fresh in the minds of researchers, the information was summarised and detailed notes were captured. Information for those notes included among other things time and date details, common themes or patterns, and any other unique observations. Researchers organised the data into different types such as, those being the observation notes, questionnaires and documents and/or artefacts. Questionnaires were also grouped according to who had completed them. This step also involved reading and re-reading the material (data) in its entirety, making notes of thoughts that sprang to mind and writing summaries of each transcript or piece of data that had been analysed. McMillan (2012:297) refers to this type of data as ‘etic data’ as they are representations of the researcher, whereas ‘emic data’ contain information provided by the participants in their own words. The objective with this step was to condense all of the information to key themes and topics that would provide some answers to the research question. The following quantitative data analysis steps were taken by researchers:

The researchers adhered to research ethics requirements. During the interactions with participants, for instance, their rights, anonymity and confidentiality were protected (McMillan & Schumacher 2014). Participants names and the specific sites they are attached to are not directly mentioned in the study. This was to further guarantee the principle of anonymity and confidentiality. Codes in qualitative data analysis were also used. Nowhere in the research was mention made of participants’ real names and the sites they worked at. During the open discussions, the purpose of the study was thoroughly discussed with the participants and they were informed of their rights to terminate or withdraw from the study, if they so wished during the process of data collection (Booyse et al. 2001). The participants voluntarily accepted the participation by signing declaration forms that guaranteed the confidentiality and anonymity (ed. Delamont 2012). The University of South Africa, College of Education Research Ethics (2015/07/15/08904782/01/MC.) granted the researchers ethical approval.

Quantitative data analysis is all about analysing number-based data (which includes categorical and numerical data) using various statistical techniques. The two main branches of statistics are descriptive statistics and inferential statistics (McMillan & Schumacher 2014).

In this study, descriptive analysis also known as descriptive analytics or descriptive statistics was used. It is the process of using statistical techniques to describe or summarise a set of data. Statements 3, 4 and 8 (Table 1 and Table 2) address the need for curriculum reform in line with the Education Act 23, 2011 and the Disability Act 6, 2012.

The statements are:

Statement 3: Has the curriculum in Zambia been reformed to facilitate inclusion of learners with SEN and LD?

Statement 4: Is government education policy on SEN understood by the teachers and other stakeholders?

Statement 8: What have been the main successes in implementation of inclusion of learners with SEN and LD in your school or centre?

The results for Statement 3 indicate that 44.6% disagreed that there had been sufficient curriculum reform as opposed to 26.6% who agreed and 28.8% opting for neutral. For Statement 4, 60.5% of the respondents agreed with the suggestion that curriculum had not been sufficiently reformed to account for the diverse needs of learners with SEN as opposed to 24.5% who disagreed with the statement. In response to Statement 8 which presents lack of curriculum reform as an impediment to the successful implementation of inclusive education for learners with SEN, 70.5% agreed with only 19.4% in disagreement and 10.1% opting for neutral. These results for Statements 4 and 8 suggest that the 28.8% neutral responses for Statement 3 might probably be leaning towards disagreement.

From the cross-tabulation results for Statements 3, 4 and 8 (Table 1 and Table 2) against all the locations, it appears that most of the respondents from the seven provinces sampled believed Zambia had seen limited curriculum reform to support the successful implementation of government policy on inclusion at the classroom level. The cross-tabulation results for Statements 3, 4 and 8 against the respondent’s current position indicate that respondents in all the current position categories generally disagreed with Statement 3, which suggests that there had been sufficient curriculum reform in line with the Education Act 23, 2011 and the Disability Act 6, 2012. The highest disagreement rate was recorded by SEN lecturers and special education needs coordinators’ (SENCOs) with 66.7% and 55.5%, respectively. In the case of Statement 4, except for the guidance counsellors and teaching assistants, all the other respondent categories overwhelmingly felt that the curriculum had not been sufficiently reformed to account for the diverse needs of learners with SEN. This result is further upheld by the responses to Statement 8 which indicate that most of the respondents agreed that the lack of curriculum reform was one of the main impediments to successful SEN policy implementation in Zambia. The average rate of agreement for all groups was 73.0%.

There were several limiting factors that had a bearing on the validity and reliability of the findings of this research study. The first factor related to the reliance on a qualitative research design which is open to subjectivity. A mixed-method approach was used to mitigate the subjectivity by including quantitative data through administering a questionnaire that yielded quantitative data and qualitative data that were analysed using a statistical software programme. In the data analysis, findings from the qualitative data and quantitative data were put together to support finding claims and inferences. Moreover, data were collected using more than three data collection tools. Consequently, validity and reliability were enhanced through triangulation of the data collection instruments in the data analysis.

The second factor related to the use of purposive sampling and the sample size. Purposive sampling might not be representative of the population and the population members do not have an equal chance of participation as would have been the case if random sampling were used. Moreover, because the participants were somewhat homogeneous as educational practitioners the data loses a critical perspective. Another key point to note about the sample is that learners with SEN were not represented as a source of data. However, it is also important to note that the nature of data required to answer the research questions made it an imperative to target practitioners who could provide relevant responses because of their experience in education provision in general, and in the provision of SEN education in particular. In mitigation of the subjectivity, the participants were targeted from different sites in both rural and urban settings. Moreover, participation was open to all practitioners as long as they gave informed consent to participate.

The third factor related to the time available for data collection which was limited to about 12 weeks. This resulted in a smaller pool of possible participants being available. Consequently, less data was collected than anticipated, which affected the reliability of the findings. Having limited time also meant that the researchers could not reach all the ten provinces of Zambia. Nevertheless, seven out of ten provinces still represent 70% reach with respect to the original intent to cover all the ten provinces in the country.

The final factor related to the emphasis on qualitative data received from interviews. It is very plausible that some respondents might have provided positive responses to avoid upsetting the status quo. However, the assurance of anonymity and confidentiality seemed to have mitigated this concern based on the patterns arising from the data analysis.

Firstly, it is imperative that the curriculum is reformed to come up with a UDL to facilitate the efficient and effective inclusion of learners with SEN in the mainstream classrooms. To this end, research-based curriculum reform should be accompanied by teacher capacity building initiatives and processes for the implementation of inclusive curriculum. It is important to note that starting at pre-knowledge in this effort and building up from there is essential for effective implementation of inclusive curriculum.

Secondly, there is a need for specific guidelines for curriculum implementation involving differentiation for the diverse needs of all learners. It is recommended that the curriculum should come with SMART objectives that are specific, measurable, achievable, realistic and are time-bound. This would allow for evaluation of inclusive curriculum implementation. Moreover, there must be clear standards for learner progress in the curriculum and that will only be possible through rigorous curriculum reform in Zambia.

Thirdly, it is recommended that there should be both support instruments and evaluative instruments at the disposal of education standard officers and other policy intermediaries to build in continuous quality assurance and quality improvement. Reflective practice is required to enhance efficient and effective inclusive curriculum implementation.

Finally, curriculum implementation will require adequate resourcing and infrastructure from government and other stakeholders in education. This calls for adequate government funding and investment in education. Construction of more classroom space, training of more special education specialists and paraprofessionals to support the inclusion of learners with SEN must go along with efforts at implementing UDL principles and practices.