New Perspectives on Maths

At the BRIDGE Maths and Science Community of Practice meeting held on 23 May 2017, Sylvester Moepya of the Harambee Youth Employment Accelerator described Harambee’s initiatives to understand, design and evaluate proxies for mathematics.

There is a pressing need to identify such proxies. Although South Africa has a vacancy rate of 20% for entry level employment, half the first-time work seekers with matric certificates who join the labour market each year do not find employment. As the NSC certificate is viewed as a poor predictor of potential, employers use other criteria for selection. Most prefer to employ people with work experience, and many require a pass in maths because they regard this as indicating learning potential and problem solving ability.  A major challenge to this approach is the small number of candidates passing National Senior Certificate (NSC) maths (only 129,481 learners in 2015).  If workable maths proxies can be identified and if these are accepted by employers, this will help young people transition into employment and will go a long way towards addressing the mismatch between demand and supply in the labour market.

Harambee has amassed considerable experience in using psychometric assessments to assess the learning potential of young people in South Africa (i.e. their ability to learn, to assimilate knowledge, and to use it functionally). A study based on 28000 individuals showed that 95% of these candidates had the required learning potential for most entry level jobs.  74% had high learning potential, but low numeracy scores. This means that employers who screen on the basis of numeracy risk missing large numbers of young people who have the potential to learn and to perform in a job.

Harambee has also worked closely with employers to understand the attributes they look for in prospective employees and to gather information on the use of maths as a proxy. This included analysing the maths needs, and the typical problems that require solving, in particular jobs (for e.g., those performed by machine operators, or cashiers). This revealed that in most entry level jobs it is mathematical literacy concepts that are required, not core maths (e.g. understanding numbers, doing conversions, manipulating data, taking measurements).

Sylvester Moepya described a particular initiative to determine whether maths was in fact a reliable predictor of learning potential and problem solving ability. This experiment was set up to test the hypothesis that candidates who had passed matric maths would demonstrate better learning potential and problem solving skills than those who had done only maths literacy, or had not done either of the two.

The experiment used a CogLab assessment tool (which measures a person’s ability to think analytically and conceptually, and to solve problems using judgement) to pre-test over 100 matriculants separated into three groups representing the three alternatives (core maths, maths literacy and neither of the two). The candidates were then divided into two groups both containing a mix of core maths, maths literacy and neither of the two. Both groups were then provided with material designed to enhance problem solving skills. One group was facilitated (i.e. given face-to-face instruction in using analytical and conceptual thinking as well as judgement to solve problems), while the second (control) group was left to proceed alone.

Surprisingly, post-testing showed that the control group (that received no facilitation) performed better than the group that had been facilitated. Could it be that people in a group led by a ‘teacher’ don’t own their own learning? Could it be that the collective energy of being part of a group where people rely on their peers and other types of problem solving helps them to perform better?

While the results for the core maths group are better, do they justify employers’ use of maths in selection?

School Subjects Sample Learning Potential Numerical Reasoning Verbal Reasoning Pre- Scores: Problem Solving – Judgement Post- Scores: Problem Solving – Judgement
Maths Literacy 38 5 27% 30% 12% 30%
Core Maths 36 6 39% 34% 20% 41%
No Maths 31 4 25% 33% 13% 32%

[Harambee is working with the Wits School of Education to take a deeper look at these results and the claims made.]

The findings on the role of maths as a proxy are particularly relevant for entry to learnerships and apprentiships. Quality Council for Trades and Occupations (QCTO) qualifications at this level require candidates to complete the Foundational Learning Competence (FLC) in both numeracy and literacy, so employers look for assurance that applicants will be able to meet this requirement. Working with an employer, Harambee has produced evidence that young people whose poor maths results or lack of maths would normally have prevented their entry to this environment, could all meet the benchmark after a targeted six-week intervention in the particular type of maths required. All the candidates were employed, and two years later are still with the company.  “The employer cannot stop singing the praises of this particular cohort,” reported Sylvester Moepya. “If the company had looked only at pre-test numeracy scores, it would never have given them a chance.”

Harambee is keen to involve practitioners and other maths education stakeholders in sharing the results of these initiatives with their own communities, and to advocate for different ways of looking at the potential of young people.