Category Archives: Uncategorized

The Comprehension Iceberg: Developing reading benchmarks in African languages


For the last few years we have been working on a project exploring reading in three African languages; isiZulu, Xitsonga and Northern Sotho (Sepedi). This was part of an ESRC-funded study looking at “Leadership for Literacy” and schools in challenging contexts. This week we published the first of a series of working papers on some of our results. The paper is titled “Investigating the comprehension iceberg: Developing empirical benchmarks for early grade reading in agglutinating African languages  (2018) and is co-authored with colleagues and friends Prof Elizabeth Pretorius (UNISA) and Mpumi Mohohlwane (DBE), who is now also doing her PhD at RESEP.

In the paper we argue here that we need to move beyond a repetitive focus on low comprehension outcomes; this is simply the tip of the iceberg. Below the surface there is widespread evidence that most children have not acquired the basic ‘tools’ for reading success – the ability to accurately and fluently decode letters and words and move from an effortful activity to an automated skill. Knowing that 78% of Grade 4’s cannot read for meaning in any language is important and sobering information, but probably not as helpful as knowing which components of reading children are struggling with and why teachers are struggling to teach them.  If children and teachers are ‘falling at the first hurdle’ (Taylor, 1989) does it make sense to focus on the seventh or eighth hurdle and ask why learners and teachers are not making it over?

One of the important contributions of the paper is that we try to estimate benchmarks for these three African languages. While many benchmarks exist in English there are currently no benchmarks in African languages and unfortunately you cannot just version the English benchmarks for the African languages because the language structures (orthographies) are so different. This is easiest to explain using an example:

Screen Shot 2018-03-13 at 11.08.10 AMSo in isiZulu the first sentence is only 3 words while in Northern Sotho (Sepedi) it’s 13 words. We go on to show that there are also large differences in ‘required fluency’ across these languages and that accuracy rates and speed differ between conjunctive and disjunctive languages – see below.

Screen Shot 2018-03-13 at 10.54.39 AM

“Comparison across the languages shows that accuracy seems to be more important for fluent reading in isiZulu than in Northern Sotho or Xitsonga. The isiZulu learners reading at 21 wcpm or faster are reading with 95% accuracy or higher. In contrast, 95% accuracy is only associated with reading at 51 wcpm or faster in Northern Sotho and 31 wcpm or faster in Xitsonga. One of the reasons why Decile-1 learners are reading so slowly is that they are making mistakes on every second or third word. The fastest Northern Sotho readers (wcpm=107) and Xitsonga readers (wcpm=91) in the sample made no mistakes whatsoever.” (p12)

This can also be seen visually:

Screen Shot 2018-03-13 at 10.55.10 AM

We then use this information and some rudimentary comprehension outcomes to benchmark the different languages:

Screen Shot 2018-03-13 at 10.54.55 AM

We’d encourage those interested in this type of work to read the full paper – if you have any comments or suggestions we’d love to hear from you! Please include them in the comments below.


The importance of learning to read in mother-tongue is widely acknowledged in the linguistics literature yet reading acquisition in African languages remains under-researched and under-theorized. While numerous studies have highlighted the low levels of comprehension among learners reading in African languages in South Africa, little has been done to understand what lies beneath this ‘comprehension iceberg.’ In this paper we present new empirical evidence on reading outcomes and the subcomponents of reading for 785 Grade 3 learners across three languages (Northern Sotho, Xitsonga and isiZulu), drawn from 61 primary schools in South Africa. This is the largest sample of such learners to date. Using an adapted EGRA-type assessment we assessed letter-sounds, single-word reading, non-word reading, oral reading fluency and oral comprehension. From this data we present results on fluency, accuracy and comprehension and how these might relate to each other in these morphologically rich agglutinating languages. We also show that there are large differences in reading subcomponents between languages with conjunctive and disjunctive orthographies. Our results suggest that there are minimum thresholds of accuracy and oral reading fluency in each language, below which it is virtually impossible to read for meaning. These are 52-66 wcpm in Northern Sotho, 39-48 wcpm in Xitsonga and 20-32 wcpm in isiZulu. We argue that there is a strong need for empirical language-specific norms and benchmarks for indigenous African languages and present our benchmarks for these three languages as a move in that direction.

“A city drowning in incompetence” – my Daily Maverick rant about the water situation in CT


(This article was first published in the Daily Maverick on the 22nd of January 2018.)

A city drowning in incompetence

This week the City of Cape Town announced that at the current rate the taps will run dry in Cape Town on the 21st of April and will soon announce the City’s 200 water collection points. These are central locations that can be secured with riot-police and include places like sports fields and schools. The Mayor has explained that after the 21st of April “citizens would line up to receive up to 25 liters of water per person, with a separate queue for the differently abled. Prior to filling their vessels, each person would be given a dose of hand sanitizer.”

I’m sorry but, what the fuck?!

How the hell did we get this close to what will be the biggest natural disaster of the post-apartheid period and the majority of Cape Townians are carrying on business-as-usual? Only 39% of residents are using less than 87l of water – the previous, and now surpassed, restriction (the current restriction is 50l person). In history, it has never happened that a city the size of Cape Town has run dry. As a Canadian headline announced this week: “Cape Town at risk of becoming the first major city in the world to run out of water” (Globe & Mail).

It is clear that there has been an outright failure of leadership in the City of Cape Town. How is it that we have been having a drought for more than 1000 days, with water management experts advising the City on a constant basis and yet only now, with 90 days to go they are escalating things? Patricia De Lille has been the mayor of Cape Town for more than 6 years and the DA has run the Western Cape for more than 7 years. It’s been years in the making. It is beyond clear that the blame for this crisis lies ultimately with the City of Cape Town and their too-little-too-late responses to an imminent catastrophe.

How has the City not already exacted drastic action to ensure that we never reach Day Zero? Why has it taken so long to radically and drastically escalate the issue in the public’s mind? The province’s ‘Resilience Officer’ dealing with the drought is currently on gardening leave because of an internal spat about the MyCiti Bus procurement processes. The media strategy of the City of Cape Town is an absolute disaster. Why has the Province not passed emergency measures in the legislature to require all Western Cape targeted media outlets to devote at least 10% of their prime-time coverage to either public service announcements on the drought or how to reduce water consumption? Why is there no blimp above the CBD announcing “Day Zero: 21 April! REDUCE WATER or taps run dry”? Instead the City pushes its pathetic #ThinkWater message. Who is advising the City on their public relations campaign?! This is about an imminent natural disaster, not a preschool slogan for Water Day. If I can get daily SMSes for insurance I don’t want, or from political parties I won’t vote for, why can’t I get SMSes about what DayZero looks like, where my local water collection point is or some equally-scary information to make me change my behavior?

Tom Eaton points out that in 2018 Mmusi Maimane has tweeted as many times about his child’s first day of school as he has about the largest natural disaster in the post-apartheid period, that is to say, once. Of the 76 tweets in 2018 by @CityofCT (their official Twitter account) less than a quarter have been about water. More than half are inane messages like “Top of the morning to you, Cape Town. With a high of 22, you can expect clouds and sunshine with a windy afternoon.” I don’t want the City to wish me a joyous morning and a jolly good night. I want them to make sure that I have water in my taps.

Why has Helen Zille not called for a localized State of Emergency, which Section 37, 1(a) of the Constitution makes express provision for. That is, that Parliament can declare a State of Emergency in the event of a “natural disaster or other public emergency.” A city of 4-million people running out of water sounds like the definition of a natural disaster to me.

Of the City of Cape Town’s 7 big initiatives to get more water online (V&A, Strandfontein, Monwabisi, Harbour, Cape Flats, Atlantis and Zandvliet), 6 are delayed and behind schedule. The City’s threats of fining excessive users or throttling their water have not materialized. Together with these empty threats and in-fighting within the DA, the City has not managed to get water pricing right or publicly shame non-compliers? Why is there no publicly available list of the top 200 addresses that are openly flouting the City’s restrictions?

It’s also not as if this is the first time this has happened in the world. From 1997 to 2009 Australia faced the Millennium Drought. Melbourne, also a city of four-million people, managed to drastically reduce water consumption by 50% and pre-emptively implement a slew of policies and programs to avoid a humiliating and dystopian scenario of a city running dry. This included setting up an emergency body mandated with decision-making during the drought. It had an expert advisory group with independent experts and full transparency (Google “Low and Grant 2015 Fighting drought with innovation” for an academic overview).

The fact that we are approaching a natural disaster and the majority of Cape Townians are still living their best lives and ignoring the extremely mild reproaches of government means that someone needs to take drastic action. Whether that is the Premier, the Mayor or the President remains to be seen. Whoever takes up that mantle needs to step into the vacuum that has been left by a wave of incompetence and denialism. Call an emergency lekgotla with all the MECs, the top 50 CEOs, the vice-chancellors of universities, editors of newspapers, the heads of civil society etc.  Pass legislation that needs to be passed, take out full-page ads in every newspaper every day. How is this not the single most important thing happening in our province at the moment?!

This is not about the ineptitude of some dingy Department of Water official failing to plan properly but about politicians and bureaucrats doing too little too late. The scale of this crisis is such that, if unresolved, it could cripple the City. A fall in tourism, a loss of jobs, declines in property prices, widespread sanitary diseases. That sounds quite dystopian, but that’s what happens when the taps run dry. The DA and the City of Cape Town need to accept that they should have escalated this issue far sooner, and then begin a massive escalation initiative to hit the complacent middle-class with real consequences. The time for kumbaya messages is over.

*Note, DM made a few edits and I chose to keep my original wording 🙂 

My Sunday Times article on Matric 2017


(This article first appeared in the Sunday Times on the 14th of January 2018).

The real matric pass rate and the real site of failure in education

Every year the matric results come out and the whole country goes ballistic. Matric is on the front page of every newspaper and swamping the radio waves and TV stations of our country. And suddenly everyone has an opinion about what’s going on in education. “The Free State is the best province”, “We should ban the IEB and have one exam”, “The rising pass rate shows our interventions are working”. No. No. No. None of this is true. Firstly, the rising pass rate – from 73% in 2016 to 75% in 2017 is purely a function of more students being held back and dropping out. In 2017 there were only 401,435 passes compared to 442,672 in 2016 – that’s a 9% decline in one year. So why were there 40,000 fewer matric passes? It’s not because the population dropped by 9% in one year (it did go down but only by about 4%). It’s not because more candidates moved to writing their matric part-time (and therefore aren’t included in the ‘official’ pass rate); there were only 11,462 additional part-timers in 2017 compared to 2016. And it’s definitely not because there are fewer progressed learners (there were 108,742 in 2016 and 107,430 in 2017, basically the same). So we know it’s not the cohort, it’s not part-timers, and it’s not progressed learners, so what is it? Basically more kids are dropping out or being held back than in 2016. It’s also no surprise that the only provinces with increasing matric pass rates (Eastern Cape +6%, Limpopo +3% and KwaZulu-Natal +6%) were also the provinces with the largest declines in the numbers writing matric (-18%, -18% and -16% respectively). So the high-level take-home is that the more you ‘cull’ and ‘gate-keep’ the more your matric pass rate increases. It also means we have 5% fewer Mathematics and 10% fewer Physical Science passes in 2017 compared to 2016.

So what is the ‘real’ matric pass-rate. A lot of people have been asking this question, including myself, and it turns out it’s a little more complicated than one thinks. Occasionally we calculate ‘throughput pass rates’ (matric passes divided by grade 10 enrolments 2 years earlier or Grade 2 enrolments 10 years earlier), yielding figures around 39% for the 2017 Grade 10 throughput pass rate. This is the incorrect figure. The problem is that at least 20% of Grade 10 enrolments are actually ‘repeaters’ and not part of the original cohort. There are also the complications of those writing supplementary exams, getting part-time passes, or Independent schools writing the NSC exams. The most authoritative research done on this is by one of my colleagues at Stellenbosch University, Dr Martin Gustafsson, and according to his calculations the ‘true’ matric pass rate is about 55% and the province with the highest achievement is Gauteng. I believe that. While this is not as bad as the crude and incorrect figure of 39%, it is also not as high as the crude and incorrect figure of 75%. If one adds in youth who get some kind of non-matric qualification (i.e. through a technical or vocational college) the figure rises marginally to about 57%. So, 43% of youth in South Africa still get no qualification whatsoever. Not matric, not TVET, not university. Nothing. They enter the labour-market ill-prepared and inherit lives of chronic poverty and sustained unemployment. Is it any wonder that since 2002 more than 40% of 18-24 year olds in South Africa are ‘NEET’; Not-Employed-or-in-Education-or-Training?

So where do the wheels come off in the education system? The answer is long before matric. When children don’t learn to read for meaning in Grades 1-3 they fall further and further behind the curriculum even as they are promoted into higher grades.  Last month the 2016 Progress in International Reading and Literacy (PIRLS) results were announced and they showed that 78% of South African Grade 4 children could not read for meaning in any language (all 11 languages were tested). And if children don’t grasp the number concept, place-value or the four operations by the time they hit Grade 4 they are on a one-way ticket to failure. The 2015 Trends in International Mathematics and Science Study (TIMSS) showed that 61% of Grade 5 students could not do basic mathematics.

I know that for those writing and passing matric it is a momentous and important occasion. It’s the culmination of our schooling cycle and (unfortunately) a rite of passage into middle class life (note: necessary but not sufficient for said passage). But as a country I feel like we need to take active steps to de-emphasize the matric results and instead focus on the ailing primary school system, the ugly step-child in our national saga. We don’t even have national assessments in primary school anymore. If the wheels come off by Grade 4 then that’s where we need to be intervening. Prevention is better than cure. A stitch in time saves nine. A matric obsessed country misprioritizes its political and economic resources.


The DBE’s 2017 NSC Exam Report can be found here.


Explaining the matric results in 7 GIFs

*11pm 5/01/18: New addendum added. Note that the throughput pass rates or comparison rates discussed here are not the ‘real’ pass rate. It is not 37% but about 55% (the former is a crude approximation and underestimates the true pass rate by about 15-18 percentage points primarily because the denominator has repeaters) See Martin Gustafsson’s comment at the end of the post. 

worked up

Every year the matric results come out and the whole country goes ballistic. Front pages of the newspapers, every radio, every TV station. And suddenly everyone has an opinion about what’s going on. All the people. so. many. opinions. “The Free State is the best province”, “We should ban the IEB and have one exam”, “The rising pass rate shows our interventions are working”. No. No. No.



  I was determined that this year I was going to stay out of the #MatricResults2017 morass and instead focus on our program to teach Foundation Phase teachers how to teach reading, because 78% of kids in SA can’t read. But no one seems to be reading the 97-page 2017 NSC Examination Report which has loads of information. Spend an hour or two analysing the data and it’s pretty easy to see what the trends are and what’s going on. So here are some of my take home points:

(1) The matric pass rate is one (very incomplete) measure of the education system and it’s a bad idea to focus on it in isolation.

got any other bad ideas (stark)

The matric pass rate is calculated as the numbers passing divided by the numbers writing. It can go up if either the numerator goes up or the denominator goes down (or both).  Nick Taylor argued this in 2011 and I’ve discussed this in 2015 and 2016. The gist of it is that it’s possible to artificially raise the matric pass rate by letting fewer weaker students write the exam (in SA this is referred to as culling or gate-keeping). So to get around this we calculate a ‘throughput pass rate’ where we divide the total number of passes by the Gr10 enrolment two years earlier or by the Grade 2 enrolment 10 years earlier. This is a VERY crude measure and does not take into account grade repetition in Gr2 and Gr10, migration, mortality etc (see Martin Gustafsson’s comment at the end of this post for a fuller discussion). So what do the trends look like in the “Traditional” matric pass rate and the “Throughput comparison rate? Basically, while the traditional pass rate has been rising the Gr10 throughput comparison rate has been declining for three years now and the Gr2 throughput comparison rate has been declining for two years. Note the ‘real’ matric pass rate is closer to 55% not 75%, but also not as low as 37%. Basically the ‘real’ pass rate is 15-18 percentage points higher than these crude estimates once you take into account repeaters and some other factors (again, see Martin’s comment at the end).

Screen Shot 2018-01-05 at 6.22.49 PM

(2) The provinces with the largest increases in matric pass rate have the largest declines in numbers writing matric

ram and cow

Basically what happens is that provinces (via principals and teachers) let fewer students reach and write matric (usually the weaker ones) which artificially lowers their denominator and raises their traditional pass rate. So the three provinces that have higher pass rates in 2017 compared to 2016 (EC, LP and KZN) also had the largest declines in the number of Gr12s writing matric between 2016 and 2017.

Screen Shot 2018-01-05 at 7.59.39 PM

(3) The Free State is NOT the best performing province in the country. Some people think that. They are wrong. 

entitled to your wrong opinion

As mentioned in point (1) above, the traditional pass rate can be very misleading. This is especially the case with the Free State. While it has the highest traditional pass rate (86%), it also has the largest decline in Gr10 throughput comparison rate between 2016 and 2017 (from 43% to 36%). What does that mean? Basically that the change between 2016 and 2017 for the Free State seems to be driven by the fact that they are letting fewer of their (weaker) Gr10’s reach matric in the first place.

Screen Shot 2018-01-05 at 6.19.30 PM

If one looks at the Grade 2 throughput comparison rate then Limpopo and the Northern Cape seem to doing quite a bit worse than they were in 2016.

Screen Shot 2018-01-05 at 6.15.27 PM

(4) So if we use the throughput pass rate who does better, the Western Cape or Gauteng?


For those who follow education statistics it’s pretty clear that the WC and GP are always the top two provinces in the country. Sometimes it’s WC (PIRLS Literacy 2016, TIMSS-N 2015), sometimes it’s GP (TIMSS-Gr9 2015). So if the Free State suddenly pops up as number 1 you’ve got to look into the data a little. And what do we find? Surprise, surprise, if you use either the Gr2 throughput comparison rate or the Gr10 throughput comparison rate then the FS is fifth. And who’s first? Gauteng and the Western Cape, kind of both. You see if you take the Gr2 throughput comparison rate then it’s GP but if you take the Gr10 throughput comparison rate then it’s WC. You usually see GP ahead if you use Gr8 throughput as well. The reason is usually because the dropout that does occur in the WC usually happens before Gr10 so their Gr10 numbers are artificially low.

Screen Shot 2018-01-05 at 5.46.42 PM

But you might then ask: why not just use the Grade 2 throughput rate instead because it’s more reflective of the whole education system? Well, it’s mainly because there is likely some inter-provincial migration between Grade 2 and Grade 12 and this is most likely from poorer provinces towards WC and GP. So the WC and GP throughput comparison rates are always slightly inflated, and the Gr2 throughput rate is the most inflated.

(5) There are other things to focus on instead of the matric pass rate

I know that for those writing and passing matric it’s a momentous and important occasion. It’s the culmination of our schooling cycle and (unfortunately for some) a rite of passage into middle class life (note: necessary but not sufficient for said passage). But as a country I feel like we need to take active steps to de-emphasise the matric results and instead focus on the ailing primary school system that seems to drop off the radar after a week of new results being released. 78% of our Grade 4 kids cannot read for meaning in any language. 66% of our Grade 9’s can’t do basic maths or basic science. Perhaps because 79% of our Grade 6 maths teachers can’t pass Grade 6 maths tests? We need to get back to basics and ensure that every child learns to read for meaning and understand the concepts of number and numeracy. If the wheels come off by Grade 4 then that’s where we need to be intervening. Prevention is better than cure. A stitch in time saves nine. A matric obsessed country misprioritizes its resources.

Transmission ends.

convo over

Addendum: Important comment from the ever-wise Martin Gustafsson (I agree with all of his points here, including that the ‘real’ matric pass rate is best estimated as 55%.

“Nic, whilst I absolutely agree with the thrust of this, I don’t agree with the 37 to 39% figure. The problem with dividing Grade 12 passes with Grade 2 enrolment 10 years ago is that there are many repeaters in Grade 2. We have relatively good data for recent years pointing to around 10% of Grade 2 learners being repeaters. We don’t know with certainty what the situation was in 2007, but the figure then was probably higher. Counting NON-repeaters only in your denominator would give you a better idea of who should proceed to Grade 12. But there are other issues: mortality, migration (e.g. across provinces if you want to look at provincial ratios), enrolment in TVET colleges, supplementary examination results, part-time Matriculants, IEB. Once you take all of this into account you get around 55%, not the 37 to 39% you refer to. The DBE has published the calculations (done by me) behind the 55%. It’s at

PS: for those who want to review the numbers and do their own calculations the Excel file and references for all the stats are available here

Open Letter to the next president of South Africa: We need a Marshall Plan for Reading

 This article was first published online on the Daily Maverick on the 6th of December 2017. It is available here
Open Letter to the next South African president: We need a Marshall Plan for Reading . – Nic Spaull & David Carel
Cyril, Nkosazana – are you sitting down?

By now we trust that your advisers have informed you about the latest reading results released yesterday: 78% of our Grade 4 students can’t read. That’s eight out of every ten children in the country. Not in English, not in their home language, not in any language. Among Setswana and Sepedi home language learners the figure is over 90%.

We came last. Of the 50 countries that participated in the test, we came dead last.

But what does that mean? After four years of full-time schooling the vast majority of our kids cannot understand what they read, if they can decode the words at all. Simple questions, workbook exercises, even the most basic story book. These are meaningless to them. And if we don’t teach our kids how to read we’ve failed them before they’ve even started.

So what does this mean to you? In all likelihood one of you is going to be our next president. Sadly, we have a fatal history of presidents denying epidemics. This is not to say that your comrades haven’t made strong statements about education – they have. They have gazetted a host of well-intentioned policies, appointed task teams, procured service providers, promulgated Norms & Standards. There has been a lot of activity.

And yet, how then do we explain that there has been no improvement in reading outcomes since 2011?

We’re not saying that your predecessors weren’t trying. Naledi Pandor implemented the National Reading Strategy in 2006 and later the Foundations for Learning campaign in 2008. Minister Angie Motshekga helped stabilise the education system with a new curriculum and is currently backing the single bright, shining star among reading interventions – the Early Grade Reading Study.

But we’re fighting forest fires with buckets. Moving the needle in education will require a radical rethinking of both what’s needed and what’s possible.

Dr Dlamini Zuma and Mr Ramaphosa, respectfully, what we do not need is another steering committee, another convening of experts, another 12-point plan. We do not need another lengthy speech or moving statement telling us how committed you are to solving the reading crisis, and to “our children’s futures”.

What South Africa needs is to decide what Japan decided in 1872, that “there must be no community with an illiterate family, nor a family with an illiterate person”. This became Japan’s ‘Fundamental Code of Education’, the core of their development strategy, and they actually did it. Within decades they had successfully eradicated illiteracy.

What South Africa needs is to decide what Cuba decided in 1961 when it implemented the Cuban Literacy Campaign. They galvanised a million Cubans to systemically eliminate illiteracy across the country. It worked.

What South Africa needs is a Marshall Plan for Reading. We need you to use yourpresidency to mobilise our country behind one goal:

That all children can read for meaning by the end of Grade 3.

When eight out of 10 of our children can’t read for meaning, overcoming this challenge might seem impossible. But insurmountable problems are not new to our country. In 2000, at the peak of the AIDS crisis, who would have thought that four million South Africans would now be on antiretrovirals? Or that the ANC government would ultimately mobilise the country to build the largest AIDS treatment programme in the world?

Will you do the same for reading?

The unfolding reading crisis: The new PIRLS 2016 results…

pirls 2

Today the PIRLS 2016 results were released by the Minister of Basic Education Ms Angie Motshekga. To say that they are anything but devastating would be a lie.  8 in 10 children cannot read for meaning. This new report provides the latest evidence helping us to understand the unfolding reading crisis. I received an embargoed copy of the final report from the IEA last week late in the evening and battled to fall asleep after reading it. 78% of South African Grade 4 children cannot read for meaning in any language. I think this was the most striking thing for me -that we had previously underestimated the number of South African children that couldn’t read for meaning. Previously we thought the number was 58% (using prePIRLS 2011 Intermediate Benchmark) but it turns out that it is 78% (PIRLS Literacy Low International Benchmark). Basically we were using the wrong benchmark in the past. This is the first time that the easier PIRLS test (which used to be called prePIRLS and is now called PIRLS Literacy) was put on the PIRLS scale.

Apart from the horrifically low levels of reading achievement, South Africa also has the highest incidence of bullying among all 50 countries that participated in the study. 42% of Gr4 students indicated that they were bullied weekly (p226 in the report). Compared to 15% in the US and England.

I’ve summarized what I think are the main findings from the PIRLS 2016 report below. You can download the full report HERE and it is also available on the PIRLS website. The SA Summary reports are now also available (SA PIRLS Literacy, ePIRLS, SA PIRLS) The DBE’s official response is here.

Main findings:
  1. 8 of 10 SA children cannot read: 78% of SA Grade 4 students cannot read for meaning. That is to say that they could not reach the Low International PIRLS Benchmark in reading. They could not locate and retrieve explicitly stated information or make straightforward inferences about events and reasons for actions (PIRLS report page 55)
  2. SA scores last in reading of 50 countries: South African Grade 4 children have scored the lowest mark in the latest 2016 round of the Progress in International Reading and Literacy Study released today. The study included mostly High Income Countries but there were a number of middle-income countries such as Iran, Chile, Morocco, and Oman.
  3. SA lags far behind other countries: While 78% of SA Grade 4 kids cannot read, in America this is only 4% and in England just 3% cannot read. However the study also included middle-income countries. In Iran only 35% of Grade 4 students could not read for meaning and in Chile it was only 13% (PIRLS report page 55).
  4. Reading crisis deeper than previously thought: When South Africa participated in prePIRLS 2011 (an easier version of PIRLS) we thought that 58% of SA Gr4 children could not read for meaning. However this was on a separate test and not on the PIRLS scale score (i.e. not the same metric). 2016 was the first time that prePIRLS (now called PIRLS Literacy) was put on the same scale score as PIRLS. The true figure for children that cannot read for meaning is 78% – revealed today. Note this does NOT mean that reading outcomes have gotten worse between 2011 and 2015. In 2011 77% could not read for meaning and in 2016 78% cannot read for meaning (this difference is not statistically significant, i.e. the difference is negligible).
  5. Some evidence of improvement in reading 2006 to 2011 but stagnant since 2011: The only good news coming out of PIRLS 2016 is that there may have been significant improvements in reading between 2006 and 2011. Because the scale scores are now comparable we can compare the performance of Gr4’s in 2006 and Gr4’s in 2011 and 2016. This comparison seems to suggest quite a significant increase in reading scores between 2006 and 2011. Notably the Gr4 students in 2011 achieved higher scores than Gr5 students in 2006. Further analysis is needed but there does seem to be legitimate evidence of improvement between 2006 and 2011. Unfortunately no evidence of improvement between 2011 and 2016.
  6. SA reading scores stagnant since 2011: There has been no improvement in reading scores over the last five years (i.e. 2011 to 2016). Note that although the average scored declined from 323 to 320 this can NOT be interpreted as a decline. The standard errors overlap here so there is no certainty that there was any decline whatsoever (this is like taking your sitting heart rate 10 times and getting very tiny differences each time – they are not statistically significantly different) (PIRLS report page 29)
  7. SA gender gap in reading 2nd highest in the world: Girls score much higher than boys in reading across the board. In Grade 4 girls are a full year of learning ahead of boys. This gender gap is the second largest among all 50 countries that participated. Only Saudi Arabia’s is higher. (PIRLS report page 36). The gap between boys and girls is also growing over time. The gap between boys and girls was larger in 2016 than in 2011 (PIRLS report page 43).
  8. SA boys scores seem to have declined between 2011 and 2016: The average Grade 4 girl in SA scored 341 in 2011 and 347 in 2016 (unlikely to be statistically significant). The average Grade 4 boy in SA scored 307 points in 2011 and 295 points in 2016 (this is likely to be statistically significant but we cannot tell until the SA report is released  (PIRLS report page 43).
  9. Declining number of SA students reaching high levels of reading achievement: In 2011 3% of SA Gr4 students reached the High International Benchmark. In 2016 only 2% reached this same benchmark  (PIRLS report page 58).

Results within South Africa:

  1. Massive provincial differences in percentage of Gr4s who can read. 91% of Grade 4 children in Limpopo cannot read for meaning with equally high percentages in the Eastern Cape (85%), Mpumalanga (83%), Gauteng (69%), Western Cape (55%). Pg 5 of this report.
  2. Very large differences by test language. 93% of Grade 4 students tested in Sepedi could not read for meaning with similarly large percentages amount Setswana (90%), Tshivenda (89%), isiXhosa (88%), Xitsonga (88%), isiZulu (87%) and isiNdebele (87%) Grade 4 learners. Pg 5 of this report.
Background: PIRLS is implemented by the Centre for Evaluation and Assessment (CEA) at Pretoria University headed by Prof Sarah Howie. CEA press release here. In 2016 it tested 12,810 Gr4 students from 293 schools across the country (PIRLS report page 309). The sample is nationally representative and can be generalized to the entire country. Students were tested in whatever language was used in that school in Grades 1-3, i.e. all 11 official languages were tested and children were generally tested in the language with which they were most familiar. The results were released by Minister Motshekga today (5 Dec) in Pretoria.
The full report is available here and also on the PIRLS website from 11am today: –
I have provided some boilerplate comment for journalists HERE. If you would like additional comment you can email me. (Comment ONLY via email. Please do not phone).

Do girls do better than boys at school & university? (my Sunday Times article)

The article below was published in the Sunday Times. The article is based on a Working Paper that my co-author (Hendrik van Broekhuizen) and I published this week. The full PDF of that article is available here.  I’ve extended that article slightly by including a few more graphs from the paper. 


Do girls do better than boys at school and at university? That’s a question that occupies the minds of many parents and teachers. The short answer is yes, they do. Many people think girls are disadvantaged in education in South Africa, and while that might be somewhat true among the very poorest girls, on average girls actually do better than boys. They learn to read much quicker than boys do (which is true of pretty much all middle- and high-income countries). In South Africa girls also perform better in mathematics. Looking at large nationally-representative surveys from 2011 and 2015 we can see that by Grade 4 girls are a full year of learning ahead of their male peers in reading, despite being in the same grade. By Grade 5 girls are about 40% of a year of a learning ahead of boys in mathematics. 

But do these advantages continue into high school and university? That’s a question that my co-author (Dr Hendrik van Broekhuizen) and I tried to address in a paper we released this week. To do this we followed the entire matric class of 2008 into and through all public universities in the country. Using data from the Higher Education Management Information System (HEMIS) we followed all the students from the matric 2008 class that went on to university and followed them for a six-year period (2009-2013). There were 112,402 students in our dataset! Because the Department of Basic Education (DBE) and the Department of Higher Education and Training (DHET) collects detailed information on the results of all students we were able to see whether or not there was also a female advantage in matric and at university. 

The results are truly remarkable. We find strong evidence of a large female advantage that continues to grow at each hurdle of the higher education process. To be specific, relative to their male counterparts we find that there were 27% more females who qualified for university, 34% more who enroll in university, 56% more who complete any undergraduate qualification and 66% more who attain a bachelor’s degree. This despite there being roughly equal numbers of boys and girls at the start of school. 

Screen Shot 2017-11-08 at 11.30.55 AM

Another striking finding is the very large drop-off from Matric to degree attainment. The diagram below shows that for every 100 females in matric there are only 85 males in matric. And for every 100 females in matric only 8 females will complete an undergraduate degree within six years, with even lower numbers for males (only 5 males).

Screen Shot 2017-11-08 at 11.31.41 AM

Because of the richness of the data we can also see whether this advantage remains after controlling for various background factors like race, age, socioeconomic status etc. We can also control for school-level achievement. We find that this large female advantage remains after controlling for school-level performance, and exists for all subgroups of race, age, socioeconomic status, and province of origin.

But perhaps it is because females choose ‘easier’ fields of study than males – and that explains the ‘advantage’? The short answer is no. We examine 19 fields of study and find that females are significantly more likely to get a degree in 12 of the 19 fields (often by substantial margins), and are significantly less likely to get a degree in five of the 19 fields. However, this is almost entirely because they do not access these traditionally ‘male’ programs rather than due to lower completion rates once they are in. Only in Engineering and Computer Science do girls do worse than boys once they are accepted to the program. But while there are these  two fields where girls do worse, there are nine where boys do worse (including Health Sciences, Business Studies, Natural Sciences, Psychology etc.). One of the most interesting findings of the paper was that females are always and everywhere 20% less likely to dropout than their male counterparts (including in traditionally ‘male’ fields like Engineering and Computer Science), even after controlling for field of study, race, age, socioeconomic status, and location.

It is important to recognize that South Africa is not an outlier in this regard. The emergence of a female advantage in education (both at school and at university) is a global trend among middle and high-income countries. In the 33 countries that make up the Organization for Economic Cooperation and Development (OECD) – mainly a club of rich countries – 58% of bachelor’s degrees were awarded to women (in SA it is 61%). The question we usually get asked is why is this the case? We don’t really know yet, but the best international evidence points to the fact that girls have more and better “non-cognitive skills.” These are things such as self-control, self-motivation, dependability, sociability, perceptions of self-worth, locus of control, time-preference and delayed gratification. Other scholars argue that schooling is set up in such a way that it favors girls over boys, and favours the traits that society expects from girls or, alternatively those that girls just naturally have more of.

Another conundrum is how it’s possible for girls to do better at school and better at university but then worse in the labour-market? Women in South Africa get paid about 15-17% less than men do for the exact same work (again this is true around the world). The answers here are also complex and link to some easily observable features of society – like who is expected to care for children – and some less observable features like patriarchal norms and gender discrimination.

What our research shows is that girls in South Africa have a clear advantage at school and at university. Any conversation about ‘gender equality’ needs to take into account the disadvantage faced by boys at school and university, but also why this reverses when one moves into the labor-market.