Advancement through the STEM fields underpins national wealth and is critical to the international competitiveness of countries. It serves as the foundation for research and development, innovation and the generation of a pool of sustainable entrepreneurs and business leaders (Harlan, 2014; Rindermann & Thompson, 2011; Tucker Drob, Cheung & Briley, 2014). Maximising STEM advancements should therefore be a critical target of educational policy (Hall, Dickerson, Batts, Kauffman & Bosse, 2011; Rindermann & Thompson, 2011; Tucker Drob et al., 2014).

However, international and national educational systems are not adequately preparing youth with the skills required to advance within the STEM fields. SA participates in three main cross-national assessments of educational achievement. In comparison to all the other participating middle-income countries, SA has the worst education system. In fact, SA performs worse than many low-income African countries in these educational achievement assessments (Spaull, 2013). Several gaps and misalignments in the SA educational system have been highlighted. Amongst these it has been identified that there is a mismatch of technical skills where the focus, level and quality of the training is inadequate (Mayer et al., 2011). Specifically, the SA Further Education and Training (FET) system has come under severe criticism over the past decade. Despite significant investments to enhance curriculum development and learner performance, the system has not achieved the anticipated output targets. This has exacerbated the increasing skills gap present in the SA potential labour market (Mayer et al., 2011; Ramdass, 2009; Smith, 2011). Rather, current educational curricula tend to be significantly more focused on verbal than on spatial abilities, resulting in a loss of students skilled in spatial ability or the reduction of achievement in this area by many individuals (Johnson & Bouchard, 2005; Kell & Lubinski, 2013).

Science, technology, engineering and mathematical-specific educational instruction has been found to influence students’ interest and achievements in STEM fields. The implementation of various forms of hands-on STEM activities was found to raise or maintain positive STEM interests amongst students at a secondary school level (Christensen, Knezek & Tyler-Wood, 2015). A comprehensive review of the empirical research available to date revealed that spatial and psychomotor training offers a promising avenue for improving student performance in the STEM fields (Adanez & Velasco, 2002; Hsi, Linn & Bell, 1997; Kell & Lubinski, 2013; Stieff & Uttal, 2015; Wai et al., 2009).

The importance of spatial and psychomotor abilities in real work settings is evident in the literature. Studies have verified that level of aptitude and skill in these areas can be influenced by situational factors such as experience, education and training (Adanez, 2004; Adanez & Velasco, 2002; Johnson & Bouchard, 2005; Kell & Lubinski, 2013; Kuncel & Hezlett, 2010). Psychological tests assessing specific cognitive areas (i.e. spatial and psychomotor aptitudes) are more effective predictors of future performance than tests that measure broader areas, such as verbal and mathematical skills. Testing candidates’ knowledge and skills within the particular course of study or career they are aiming to pursue is likely to result in the greatest return on investment as more ability in these areas results in enhanced performance (Kuncel & Hezlett, 2010; Watkins, Lei & Canivez, 2007).

The value of psychological aptitude tests therefore lies in their ability to successfully identify individuals with potential and those who are trainable in specific abilities (Foxcroft & Roodt, 2013; Furnham, 2008; Ter Laak, Gokhale & Desai, 2013). Cognitive and psychomotor tests have been found to be useful in employee selection for a wide variety of occupations, mostly being part of multiple-aptitude batteries (Carretta & Ree, 2000). Research done by Hunter and Hunter (1984) revealed that psychomotor ability had greater incremental validity than general cognitive ability for less cognitively complex jobs, such as elementary industrial workers and vehicle operators. The import of assessing spatial ability was also highlighted in a longitudinal study, which revealed that spatial visualisation had a lower correlation with socio-economic status, in comparison to mathematical and verbal reasoning abilities (Shea, Lubinski & Benbow, 2001). A compelling case for the logic in including spatial aptitude assessments in selection exercises for STEM-related disciplines is put forward by Kell and Lubinski (2013). However, the vast majority of studies investigating psychological tests have pursued an investigation of the utility of using general cognitive ability tests to predict educational and vocational outcomes (Foxcroft & Roodt, 2013; Furnham, 2008; Ter Laak et al., 2013). There is currently a lack of research on the impact of specific cognitive abilities, such as spatial and psychomotor aptitudes. Most of the research on spatial ability has been on its biological underpinnings. Therefore, there is a call for research to be done into the role of spatial and psychomotor ability on selection as well as the effect curriculum and training has on these aptitudes (Kell & Lubinski, 2013; Shea et al., 2001). Very little research has been done on the secondary education factors that impact on spatial cognition and on the potential mechanisms that can be employed to improve these malleable skills (Stieff, Dixon, Ryu, Kumi & Hegarty, 2013; Uttal et al., 2013a; Uttal, Miller & Newcombe, 2013b).

It is with this in mind that this study sought to investigate the relationship between the technically oriented aptitudes and the forms of SA secondary education (incorporating subject choice). The secondary objectives of the study were stated as follows:

To determine whether there is a relationship between level of eye–hand coordination and type of secondary education (incorporating subject choice) attained;

In order to address and expand on the abovementioned research objectives, the five primary research questions underpinning the project were:

Is there a significant relationship between the type of secondary education (incorporating subject choice) obtained and hand–eye coordination ability?

In the section that follows, a literature review of the theoretical frameworks underpinning the study is provided alongside a review of the empirical literature in the research field.

The structure and mechanisms of human intelligence have been researched for more than 100 years (Foxcroft & Roodt, 2013; Furnham, 2008; Nisbett, 2013). A dominant model on the structure of human intelligence is the theory of fluid and crystallised (Gf-Gc) intelligence, initially developed by Cattell (1963) and developed further by Horn (1976). However, because of various criticisms, there has been a move away from this viewpoint of general intelligence (Ekinci, 2014; Johnson & Bouchard, 2005; Nisbett, 2013; Shahzada, Khan, Ghazi & Hayat, 2015).

Notably, a multiple perspective of intelligence was proposed by Vernon (1964, 1965, 1969) and Humphreys (1962), who did not adhere to the commonly held belief of the time that the g-factor was a primary and important determinant of intelligence. Whilst Vernon (1969) upheld the importance of general intelligence, he did highlight that once the general intelligence factor was extracted from any compilation of ability tests, the residuals fell into two main categories. The first of these is labelled v:ed, referring to verbal and educational skills that assess verbal fluency, verbal and numerical scholastic knowledge as well as divergent thinking. The second residual is referred to as k:m and unpacks spatial, practical and mechanical abilities, including perceptual speed and psychomotor abilities (Johnson & Bouchard, 2005; Vernon, 1964, 1965, 1969). Empirical support and the practical importance of these two residual grouping of abilities were found in the prediction of work performance of sizeable military samples (Humphreys, 1962).

The focus of this study was on aspects of Vernon’s (1969) k:m intelligence and two forms of Gardner’s (1999) multiple intelligences, namely, bodily kinaesthetic intelligence (KI) and visual and spatial intelligence (SI). These specific cognitive abilities or aptitudes are narrower than general cognitive ability and tap into different hierarchical levels of human intelligence (Carretta & Ree, 2000; Johnson & Bouchard, 2005; Johnson, Nijenhuis & Bouchard, 2007; Rohde & Thompson, 2007). KI is an individual’s capacity to skilfully utilise and control their body in terms of psychomotor activities. The effective employment of equipment and machinery are also expressions of this form of intelligence (Gardner, 1999). Also termed psychomotor ability, this form of intelligence measures skills such as finger dexterity, control precision, multi-limb coordination and rate control (Carretta & Ree, 1997; Chaiken, Kyllonen & Tirre, 2000). This study explored two types of psychomotor abilities: two-hand–eye coordination and one-hand–eye coordination in the first and second research questions. Spatial ability or SI is the capacity to visually perceive sensory data with the mind’s eye and to accurately appreciate the technicalities of space. The ability to mentally rotate or manipulate this sensory input in space or given area is also included in this specific aptitude (Carroll, 1993). In this study, the third to fifth research questions investigate this form of specific aptitude through testing the applicants’ ability to mentally assemble representations as well as their 2-D and 3-D spatial perception.

The different forms of intelligence appear to be differentially affected by environmental factors such as age, education and training (Carretta & Ree, 2000; Watkins et al., 2007; Schmidt & Hunter, 2004). International research has confirmed that spatial ability is an aptitude that can be improved with training (Adanez & Velasco, 2002; Hsi et al., 1997). A large longitudinal study investigated the impact of mathematical, spatial and verbal assessments on a range of educational–vocational outcomes amongst high school students. An interesting corollary from this longitudinal study was that spatial ability has a lower correlation with socio-economic status (SES), relative to both mathematical and verbal reasoning abilities. More spatially intelligent individuals from lower SES levels could therefore be identified through the use of spatial ability measures (Humphreys, Lubinski & Yao, 1993). Furthermore, a recent meta-analysis of 217 research studies established that spatial ability can reliably be improved through training. This meta-analysis investigated several moderators, including age, gender and type of training and found the findings to be stable across these moderators. Training effects were also confirmed to be durable and generalisable (Uttal et al., 2013a). This has also been supported at a tertiary level. Pre- versus post-visualisation test scores of first-year engineering students revealed that the teaching methodologies employed in developing this aptitude resulted in more than a third of the class increasing their performance on the test (Adanez, 2004). Training in spatial ability has been shown to be both durable and transferable. The impact of the training was observed months after the workshops and were being applied in tasks at least moderately different to those experienced in the training (Uttal et al., 2013b). Research conducted amongst first-year engineering students in Brazil highlighted that the early detection of students with low spatial aptitude, via a visualisation psychometric test, was able to improve these students’ performance levels in technical drawing. The researchers also held that the elementary and high school activities in use at the time of the study did not adequately promote the development of spatial aptitudes within these Brazilian students (Adanez & Velasco, 2002).

Currently, there is significant appreciation of spatial ability in general terms. However, there is still a gap in the literature pertaining to the attainment and development of SI in educational settings and the impact thereof on the occupational field (Johnson & Bouchard, 2005; Kell & Lubinski, 2013). In the range of multiple intelligences or the repertoire of human cognitive abilities present in human beings, SI needs to be explored further. This is especially necessary given the established link between this attribute and performance in the STEM fields (Kell & Lubinski, 2013; Shea et al., 2001; Uttal et al., 2013a, 2013b).

Psychomotor and dexterity testing have also been increasingly used within the human resource management context (Foxcroft & Roodt, 2013; Furnham, 2008). Psychomotor tests are readily employed as employee selection measures for apparel manufacturers. However, one study of sewing machine operators, in eight different plants, failed to show a significant relationship between the three psychomotor test scores and subsequent measured productivity (Inskeep, 1971). Psychomotor and dexterity testing have been successfully employed in the selection of students for surgical training (Gallagher, Leonard & Traynor, 2009). Two spatial visualisation psychological tests, two-hand coordination and complex coordination tests were also used in the selection of pilots as part of a multiple-hurdles method (Bordelon & Kantor, 1986). A manual dexterity test was furthermore employed to predict students’ grades on courses at a pre-dental school to ascertain whether the students were ready to advance to the clinic. This test did not appear to add information to the current admissions criteria (Gansky et al., 2004).

The SA manufacturing sector requires high-level competencies in order to keep abreast of global competitiveness, the associated technological improvements and a significant increase in the prevalence of lean manufacturing practices (Mahembe, 2012; Nzimande & Patel, 2012; Rauner, Heinemann, Hauschildt & Piening, 2012; Shatouri et al., 2012). Well-developed practical and technical skills are therefore recognised as being vital in entry-level technician recruitment (Phillips, 2003). Empirical research conducted by the merSETA highlighted that 76% of the manufacturing sector’s employers regarded these skills as fundamental in their recruitment of artisans (Peo, 2013).

The dilemma within the SA context is the selection hindrance created by the current inefficacy of the SA secondary education system (Deloitte, 2009). Education is the main cause of elevated skills and hence improved performance and productivity within a knowledge economy (Akinyemi & Abiddin, 2013). However, a national survey investigating the main factors limiting the success of businesses in South Africa identified an insufficiently educated workforce as one of these determinants (Nzimande & Patel, 2012). The achievement of a Grade 12 qualification should be a useful gauge for the measurement of skill levels in a country. However, it has been argued that the SA education system may be a deficient preparation base for employment and that the achievement of a secondary education may not be a dependable indication of learning capability (Cosser & Sehlola, 2009; Lehohla, 2010; National Treasury, 2011).

The SA automobile industry has traditionally centred its recruitment and selection of entry-level employees on individuals with a technical education (Schafmeister, 2013). However, the number of learners who are graduating with technical Grade 12 qualifications that support entry and success at the higher educational level is not increasing proportionate to the increased demand for these skills (Mahembe, 2012; Nzimande & Patel, 2012; Rauner et al., 2012). The merSETA has recently recommended a 50% or more pass in both mathematics and science for apprenticeship selection (Brown, 2013). In order to enhance the supply of adequately educated learners to the manufacturing sector, the merSETA has launched a number of projects that aim to improve Grade 12 mathematics and science performance (Patel, 2012; Peo, 2013). It is the intensified focus on the inclusion of these two subjects within the selection practices employed within the manufacturing sector that led to the study’s five hypotheses:

Hypothesis 1: There is a significant relationship between levels of dexterity and type of secondary education (incorporating subject choice) obtained.

For ease of reference, secondary education is understood as collectively referring to technical high schools, academic high schools and FET institutions. The terms ‘secondary education’, ‘Grade 12 qualification’ and ‘matriculation’ are hence used interchangeably. In this article, five forms of secondary education are named and defined as follows:

A Grade 12 education with mathematics and science as subjects in a secondary education qualification obtained through an academic high school.

The quantitative research design and methodology utilised is outlined in the section that follows, as well as the study’s findings. Based on these findings, conclusions, recommendations and limitations of the research are provided.

A quantitative research methodology was followed to achieve the research objectives using a cross-sectional design. With a cross-sectional design, numerous individuals are examined at a specific point in time (Bryman, 2012; Neuman, 2012). This approach is appropriate for this study as it was economical and time effective. The study is descriptive in nature as the study aimed to investigate whether a relationship exists between secondary education (incorporating subject choice), eye–hand coordination and spatial abilities (Jankowicz, 2005). The sample that the research targeted was all the applicants who successfully completed the general aptitude phase in the selection process for operators within an automotive plant in the Eastern Cape. Mathematical literacy was not a subject investigated in this study. The type of sampling method used was convenience sampling, which involves selecting a sample according to the availability to the researcher (Quinlan, 2011). The data analysis was done by making use of the Chi-square test of independence to ascertain the relationship between the variables.

Table 1 displays the descriptive statistics for the five aptitude sub-tests. The 2-D and 3-D spatial perception tools recorded the highest medians at seven, being in the high range. The medians for the assembly and coordination sub-tests were within the moderate range, being six and five, respectively. The lowest median, three, was found for the dexterity measure. The highest mode, nine, was established for the spatial perception 2-D sub-test, with the second highest mode, being seven, recorded for the spatial perception 3-D instrument. The mode for the assembly sub-test was six, whilst a mode of five was documented for the coordination sub-test. The mode for the dexterity sub-test was the lowest at three.

Chi-square tests of independence were completed to scrutinise the relationship between the form of secondary education (incorporating subject choice) and the stanine scores achieved in the five aptitude tests, grouped according to low, moderate and high scores. Tables 2 through 6 provide a synopsis of the results obtained by the applicants (N = 1566) in the five aptitude measures according to the form of secondary education obtained (incorporating subject choice).

Three conclusions have been derived from the study’s five hypotheses. The first conclusion points to a differing complexity level across the five instruments employed in this study, whilst the second and third conclusions outline the different levels of aptitudes presented by matriculants with different types of secondary education.

In Table 1, which summarised the distribution of the five aptitude sub-tests, the increased intensity of exertion is clearly evident from the wide frequency distribution across the stanine scores of the five tools. The median and mode for the dexterity measure, which assessed the respondents’ two-hand–eye coordination, is significantly below those of the other four tools. These two central tendency measures place the results for this sub-test in the below-average category, indicating that the majority of the respondents would be outperformed by at least 77% of the SA population in this aptitude. The median and mode for the one-hand–eye coordination instrument was slightly higher than that obtained for the dexterity sub-test. The three measures of spatial visualisation realised central tendency stanine scores in the high-average to superior range. Thus, it is concluded that the respondents found the two psychomotor tools more challenging than the three spatial visualisation techniques. Previous research has offered two explanations for this lower ability on high-speed aptitude tests. Firstly, because of outcomes-based education, SA students tend to be less speed oriented (Van der Merwe, 2008). Secondly, the maintenance of selective attention in the youth has been radically reduced because of the continuous over-stimulation offered by readily available entertainment (Cocodia et al., 2003).

The second conclusion pertains to the endorsed types of secondary education. Based on the results obtained across the five aptitude measures, three forms of secondary education (incorporating subject choice) are recommended, namely a Grade 12 qualification inclusive of science; a matriculation education with mathematics as a subject; and a secondary education including both mathematics and science subjects. The applicants with the secondary education inclusive of science consistently outperformed the other applicants educated through matriculation type in three of the instruments, namely, dexterity, assemble and spatial perception 3-D. However, in two of the sub-tests, namely coordination and spatial perception 2-D, they were outperformed. Matriculants who studied mathematics and/or science performed consistently better in the five aptitude sub-tests in comparison to the other two types of secondary education. These three forms of secondary education are concluded to have been positively associated with the applicants’ accomplishments in these specific aptitude measures. These forms of Grade 12 qualifications may also have afforded the applicants with exposure to a spatially enriched education as described in the research by Hsi et al. (1997), Stieff et al. (2013) and Uttal et al. (2013a). This conclusion is in line with previous research which established that enhanced spatial ability is significantly associated with accomplishment in STEM domains (Humphreys et al., 1993; Kell & Lubinski, 2013; Shea et al., 2001).

Thirdly, the technical and general types of secondary education were not optimal performers in the five aptitude sub-tests. Whilst the technical matriculation applicants performed moderately well in the assembly and spatial perception 2-D sub-tests, they did not perform optimally in the other three sub-tests. It is therefore concluded that these two forms of matriculation (incorporating subject choice) education are not able to reliably provide the technically oriented abilities required of automobile operators. This confirms the abundance of literature indicating the educational crisis that is being experienced in the technical school system, both internationally and nationally. Improved technical training systems should assist in addressing the current technical skills gaps by focusing on the type, relevance and levels of courses being offered (Ramdass, 2009; Spaull, 2013; Van der Berg, Taylor, Gustafsson, Spaull & Armstrong, 2011; Van de Werfhorst, 2014). This study has emphasised the challenges faced by national FET institutions (Department: Higher Education and Training, 2014; Manuel, 2011; Nzimande, 2010). Whilst a high demand for learning opportunities exists within the SA youth, FET colleges have not been successful in enticing this portion of the population to participate in their programmes. This may be because of the weak relationships between FET colleges and the workplace, which culminates in unsuitable types and levels of training (Manuel, 2011).

Overall, the results of this study have highlighted statistically significant relationships between the five technically oriented aptitudes assessed in this study and the type of secondary education (incorporating subject choice) attained. This finding supports previous research that has highlighted the need for changes in curriculum design, teaching and assessment methodologies currently being employed in certain forms of secondary education. Being exposed, through certain forms of schooling, to specific stimuli and activities may have significant associations with specific abilities and hence potential employability.

The contribution of this research study is three-fold. The study has confirmed the applicability and relevance of employing the type of secondary education (incorporating subject choice) obtained as a cost-effective screening tool. Human resource practitioners need to take cognisance of the warning provided by several academics and researchers regarding the poor quality of education stemming from certain forms of secondary education in the country (Ramdass, 2009; Spaull, 2013; Van der Berg et al., 2011; Van de Werfhorst, 2014). Technical high schools and FET institutions need to review their current offerings in order to equip students adequately with the specific aptitudes required within the automotive industry. Candidates with a technical-type secondary education performed inadequately in the two psychomotor and spatial perception 3-D sub-tests in comparison to the other four forms of matriculation. Technical training and education providers need to address these specific areas in their future offerings. Open discussions with the broader manufacturing sector may assist in improving understanding of the industry’s requirements.

Secondly, the study has provided SA research confirming a relationship between education, training and specific cognitive abilities. This is of particular importance given the documented association between spatial aptitudes and achievement in STEM fields. Thirdly, this study’s findings have highlighted the benefits of utilising a multiple rather than a singular selection methodology. This is highly relevant in the SA context where supply exceeds demand in the labour market.

This study’s sample is not in line with either provincial or national data pertaining to age, race or gender as indicated in the 2001 census (Lehohla, 2004). Furthermore, this study was part of a sizeable recruitment process implemented by a SA automotive assembly plant with the sampling frame for this study being a database of pre-screened potential operators. Therefore, the main inadequacy of this research is that the results cannot be generalised to either the SA or global automobile industry. The second limitation of the study relates to the scores obtained by the applicants in their secondary education qualifications, both overall and for specific subjects. This was delineated out of the study’s scope but it is plausible that the sample constituting the technical and general types of secondary education was not of an adequate calibre. This should be explored in future research. It is also recommended that future studies explore the relationship between the types of secondary education and employability as conducted by Shea et al. (2001) and/or on-the-job performance ratings.