Paul Baker: Understanding the gender gap - part 1

Dr Paul Baker, Rector of Waitaki Boys' High School and a member of the Ministerial Reference Group set up two years ago to guide the Government on boys' education is speaking today to a conference on Boys' Education at Massey University Albany Campus. These are his speaking notes.

A four-year study observed how 14 young chimps in Tanzania learned, by observing their mothers, how to catch termites, using a thick stick as a tool. The mothers showed no gender preference in teaching. The daughters copied them closely while the sons "would quickly lose patience and play games". On average, the females learned the skill in 31 months; the males 58 months.

Adrienne Alton-Lee cautions about using the gender gap "as a kind of absolute measure of what matters". But, even in the Simian world, it's the gender gap by which we have come to judge male achievement - and in the human world, it is changes in the gender gap, more than any other factor, that brings us here today.

Good afternoon, everyone, and a special greeting to Team Waitaki, and to my wife who is here to ensure that this speech is finally delivered and over, so the gender gap in our domestic duties can be diminished.

My aim today is to advance our understanding of the gender gap in New Zealand. Much of this presentation has had to be researched from scratch. The research base on boys' education in New Zealand is minimal. In consequence, this might be entitled the Hans Brinker Memorial Speech, as I'll be skating on thin ice throughout. In various aspects of the topic I've done enough research to form a hypothesis, but not enough properly to test it. I'll be identifying ten areas where further research is vital, if we are to do our best for the boys of this country.

I'll be arguing the following

1. The Level 1 gender gap is significant and has existed since 1993.

2. The Level 3 gap is larger than Level 1, larger than normally reported, and still growing.

3. The gender gap is subject based, and influenced by how we choose to construct subjects, teach and assess them.

4. The gap reflects and may contribute to changing gender patterns in tertiary education and employment.

5. The gap is not connected to race, class or rurality.

6. There is a crisis in the gender composition of the primary work force and a potential crisis at secondary level.

7. Boys' schools or classes have particular advantages in meeting male needs.

8. Most co-ed schools are doing little to specifically target boys' needs, but boys can thrive in co-ed schools that are well run.

9. There are no simple solutions, but there are complex, multi-layered solutions.

10. New Zealand's institutional response to the gap has been one of denial, delay and trivialisation.

The gender gap starts in the womb. But we'll move quickly on, to the start of schooling, where entry testing shows girls much better prepared than boys.

By years Four and Five we find

* No significant differences in most subjects, but
* Distinct gender preferences within subjects (i.e. physical science for boys, life science for girls)
* Female superiority in writing, speaking and reading.

In two international tests New Zealand had the 4th and 2nd largest gender gaps in reading.
* In NEMP testing Year 4 Girls performed better at 14 per cent of tasks, and boys at 8 per cent.

At Year 8, Girls performed better at 22 per cent of tasks and boys at 6 per cent. The gender gap grew hugely in writing. With that exception, NEMP directors Terry Crooks and Lester Flockton don't consider the gap statistically significant. They call the moral panic about male achievement at primary level unwarranted.

National and international testing still shows similar gender performance in Year 9 (Science and Maths) and even in Year 10, with boys ahead in Reading Comprehension, Reading Vocabulary and Mathematics.

Gender Gap in Level Completion

It's from Year 11, and for external qualifications that now reflect motivation and work ethic, as well as just ability, that a substantial and comprehensive gender gap emerges, of around 10 percentage points. (I regret that, unless otherwise stated in this paper, gender gap means a gap favouring girls. It's usually expressed as the percentage points; i.e. a female pass rate of 60 per cent and male rate of 50 per cent gives a gender gap of 10 percentage points).

NCEA Level 1

External achievement standard results show that the gap spans all subjects, but at a level we would consider problematic only in Languages, English and Arts (where the gap is dropping a little) and Technology (where it is sharply rising).

NCEA Level 2

At each level, Maths, Science, and Social Sciences had gaps of only 1-3 per cent.

NCEA Level 3

They account for the two thirds of external results. If they accounted for three thirds of results, we would not be here now.

Their small gaps are significant only because these subjects once generally favoured males and thus statistically balanced the male deficiencies in other subjects.

Although the gender gap varies in size, it is comprehensive. Of 296 external achievement standards, girls outperformed boys in 258, mainly in culture or performing arts, literacy, human biology, health, social issues, and in technology, where six standards, all with large and gender-balanced enrolments, had gender gaps ranging from 22 per cent to 43 per cent.

Boys outperformed girls in 38 standards: mainly relating to economics, statistics or production.

To conclude: we appear to have a stable gender gap of around 10 percentage points, spread across all subjects and most standards, but significant only in some.

How serious is this?

If we express it in every day terms, it means that for every six girls who pass, only five boys do.

Gender and Ethnic Gaps

If we express it in comparative terms, it's much smaller than ethnic gaps of 22 and 25, while the socio-economic gap, between low and high deciles, is 30.

It seems that gender must take its place in the queue.

But wait. NCEA pass rates of each gender are the common basis for understanding the gender gap. It's the 10 per cent that generates the headlines. But pass rates suppress the real gender gap. This is because they are the pass rates only of those still at school.

Students leaving Before Year 13

But boys have a lower retention rate than girls. By the end of Year 12, 43 per cent of boys but only 35 per cent of girls have left school.

When a lower male enrolment is combined with a lower achievement rates of those who are enrolled, the gender gap seems much larger.

Males as per cent of enrolled students and students passing

In 2005 males accounted for 48 per cent of the Year 12 roll and only 44 per cent of Level 2 passes.

They accounted for 47 per cent of the Year 13 roll and just 41 per cent of Level 3 passes.

Because of falling male retention, the gender gaps at Levels 2, and particularly 3, are much larger than commonly understood and reported.

At Excellence level the gap is even greater. In 2005 girls gained 75 per cent more excellence passes at Level 1 and 57 per cent more in Level 2 and there were 50 per cent more female Outstanding Scholars.

Let's now look at when and why the gender gap emerged.

According to Shakespeare, it was a very long time ago.

The whining schoolboy
with his satchel
And shining morning face
Creeps unwillingly to school
(As You Like It, 11, vii)

Three centuries later a British Schools Inquiry effectively wrote the same thing: "Girls come to you to learn; boys have to be driven". Male apologists in fact considered "healthy idleness" to be a defining quality of masculinity. A studious male might be regarded with suspicion - two educators wrote in 1913 "The boys' breezy attitude to life…successfully secures him from morbid concentration on the acquisition of knowledge". Like whites living under apartheid, what did it matter if males were poorly educated? The privileges of society were still theirs by right.

The earliest year I have New Zealand data for is 1970, so let's start then, my School Certificate year. These were the gender issues: boys dominated the classroom and leadership positions, and harassed girls who were passive and not keen on Maths and Science, which limited their career opportunities.

What no-one seemed to find problematic was that just ten girls throughout the country wanted or were allowed to take Mechanics, engineering, woodwork, workshop technology and agriculture and just eight boys had penetrated clothing and textiles, homecraft, and shorthand typing. Thirty were doing typing, and, by a quirk of timetabling that denied us any alternatives, they included myself and five male classmates. We had to wear bibs so we wouldn't look at the keys, an extreme humiliation whenever messengers entered the room, and definitely not boy-friendly. Some of us managed modest School Certificate passes.

Gender Gap in SC Subject Passes 1970-1993

This slide shows how today's gender gap emerged. In 1970 there was already a small gap, of 2.3 Girls led in 12 subjects and boys in nine.

Ten years later the gap had increased slightly to 3.3. The subject ratio was unchanged.

The next year I have figures for is 1987. Four School Certificate grades had become seven, so we're not making precise comparisons, which may explain why the gap has now dropped to 1.6. Girls ahead in 13 subjects, boys in 12.

In 1989 the gap is 2.7, perhaps evidence of an emerging trend.

Frustratingly, NZQA has no record of subject passes by gender for the crucial years of 1990 and 1991.

By 1992 seven School Certificate grades had become five, so again, direct comparisons are not possible. But for 'pass' grades the gender gap was now 4.9. In 1993 it was 6.1. Boys were ahead in only four of 21 subjects.

Gender gap in SC/NCEA

This slide shows how consistent the gap has been, once established.

It peaked in 1996 at 6.8, and then varied between 4.4 and 6.4.

The gap in top grades followed a similar pattern.

NCEA made little difference, especially to top grades. Individual subject pass rates across the two systems of School Certificate and NCEA also show remarkable consistency.

Why then was NCEA believed to have increased the gender gap?

Because School Certificate did not have an overall pass rate, and NCEA does, statistical apples were compared with statistical lemons. The reported jump was from a 5-6 point School Certificate gap (taken from individual subject passes) to a 10 point NCEA gap (taken from overall level completion passes). Level completion is harder than passing one subjects, girls work more consistently across subjects, and enter more papers, so the level completion gap will be greater than the subject completion gap.

Our current Level 1 gender gap, which some still consider news, was well established by 1993. There's been no real change since then; only a change in the method of reporting.

Analysing what happened between 1990 and 1993 might shed more light on the gap today.

The gap increased in every subject except French. However:

It increased the least – or not at all – in subjects where boys traditionally did badly: Art, Languages, and particularly in English where there has, quite remarkably, been no movement at all from 1970 to the present.

In Maths the gender gap was 2 in 1970 and still 2 in 1990.

The swings were biggest in subjects where boys traditionally did well, notably Physics, Economics and Accounting. In these subjects probably lay the greatest potential for female improvement, for teacher encouragement and efforts towards gender inclusivity.

Girls can do anything. Is that a sufficient explanation of the dramatic rise of the gender gap in the early 1990's.

I've struggled to identify any other obvious factor.

How we teach can be divided into curriculum, pedagogy and assessment, and none of these areas is necessarily gender-neutral. But, other than Economics adding internal assessment, nothing specific happened in these areas during the crucial years of 1990-3.

After then, however, developments in curriculum, pedagogy and assessment all favoured females, and thus maintained a gender gap that might otherwise have dropped.

Firstly, curriculum. Even within subjects genders have preferences. Look for example at the varying gender gap for English level 1 achievement standards, boys doing best in the standard that requires least preparation. Many subjects have some standards in which girls do best and some favouring boys.

English 2004 Level One Externals 

Therefore how we choose to construct subjects – what we put in and what we leave out – can have profound gender implications. For example

In 1970
* The subject Science/Physics had a gap of 11, while the subject Physics had a gap of 6 favouring males.
* General Science had a gap of 6 while new Science Syllabus favoured boys by 8

In 1980
* The Science gap was 2 but Alternative Science 6.
* Four different Maths Internally Assessed Papers, gave gender gaps ranging from 11 to boys, to 3 to girls.

New or growth subjects of the 1990's - Agriculture, Horticulture, Human Biology, Japanese – all favoured girls. So did the reconstruction of traditionally male technical subjects: in 1994 Tech Drawing became Graphics and the gap went from an average of .6 to males, to 10 to females. In 1998 Workshop Technology became Design Technology and the gap went from an average of 12, to 17.

By 2001 gender apartheid was dead - I surveyed co-ed Principals on this issue and they overwhelmingly felt that subjects are no longer gender typed - 4,500 girls were taking traditionally maletechnical subjects and in equal number of boys were taking Home Economics, Human Biology, and Text and Information Management. The difference: boys were doing badly in them. When in 1998 Home Economics became Food and Nutrition, boys did even worse.

How we teach – pedagogy – is also not gender free. There has been a shift from closed, structured, information-dense learning activities, which boys did better at, to open-ended, experiential, reflective activities. Ian Lillico considers this an example of the "effeminized curriculum".

Traditional subjects have become more literate. In subjects like Technology and Physical Education, many boys just want to "make things" or "do things". They do not particularly want to plan, or write about their plan, or keep a log, or engage in post-modern deconstruction analysis.

Success in assessment now requires understanding and meticulously meeting complex written instructions. This clearly favours girls, because of their superior language and organizational skills, and ability to understand and deliver what others expect of them, which many teenage boys have never been too bothered about. This was nicely encapsulated in a recent article interviewing senior Palmerston North students on NCEA. One Anna Neld liked it as students now knew "exactly what they need to do", while James Benn, speaking for the male team, called it a "bit of a pedantic system" where you could lose marks for small errors.

Here's an example of how assessment is not necessarily gender neutral. Before 1993, Bursary Agriculture favoured girls. In 1993 there was a new examiners team. The gender gap abruptly reversed. In 1997 it abruptly reversed again.

A more recent English example shows how assessment change can deliberately favour one gender. In 1999, when low male test scores became a political issue, the age 11 reading test was made more 'boy-friendly'. It involved three short passages about spiders, and most marks were given for factual comprehension. Boys' reading scores leapt by 14 percentage points, just 6 points behind girls. There was further controversy in 2001 when the reading tests were "fact-filled, non-fiction text in magazine format, chopped into bite-sized chunks". Boys' scores improved again.

I recently surveyed co-ed principals on a number of gender issues. Many of the changes to curriculum, peadgogy and assessment that seem to benefit girls of course pre-dated NCEA, but NCEA is now seen to represent the whole package. I asked whether NCEA favoured girls, or boys, or was neutral. Two thirds of respondents considered NCEA favours girls. One third considered it gender neutral. Not one principal considered that NCEA favours boys.

Changes to the way we construct, teach and assess subjects clearly did not cause the gender gap. But they are now contributing to it.

Few would argue that boys are more literate or innately creative than girls, and therefore some sort of gap in English and creative subjects is expected. But if a subject such as Phys Ed or Technology attracts large numbers from both genders but girls consistently perform better, do boys need to work harder, or does the construction of the subject merely reflect the tertiary or employment requirements related to that subject – that may also leave boys at a disadvantage - or is the subject incorrectly constructed or assessed? More research and debate required.

The gender gap at Level 3 has a different dynamic and requires separate consideration. It is larger than the Level 1 gap, and growing.

The gender gap in the award of Bursary was 4.2 (2001) 3 (2002) and 1.9 (2003). The gap in NCEA Level 3 jumped to 10.7 (2004) and 10.0 (2005).

A similar picture is shown by expressing the males who passed as a percentage of all passes. This figure has dropped from 53 per cent in 1988 to 41 per cent now.

Male Bursary/Level 3 passes as a percentage of all passes.

NCEA doubled the gap because Level 3 was easier than Bursary. The number of males who gained it increased by 60 per cent over the previous year's Bursars, but the female increase was 78 per cent.

In 1992 females were ahead in 17 Bursary subjects and males in 8. By 2003 females were ahead in 22, and males in just three, all by under 2 points. Only in Accounting, were males consistently ahead 1992-2002. No subject showed a dramatic gender reversal, nor were there any new or reconstituted subjects; just steady attrition.

Proportion of University Students aged 17-20 who are male

The gender gap is growing at Level 3 because it is growing at tertiary level. As a proportion of University students aged 17 to 20 males have dropped in just five years from 46 per cent to 42 per cent. This appears to be the most dynamic trend of any gender gap, and its quite international.

The tertiary trend in turn reflects - and possibly causes - changes to employment patterns. Particularly in the current economic climate, where skilled tradespeople are well paid and sought after, males seem keener than females to start work, and in jobs that may require only Level 1 or 2. Women continue to shun such jobs - in building, motor mechanics, boatbuilding and plumbing women they are just 1 per cent of modern apprentices – and seem more prepared to continue studying, and the jobs they want are more likely to require tertiary qualifications. As we move from a productive to knowledge and service-based economy, employment opportunities for people who have good inter-personal relationships, responsiveness to client demand, and high levels of literacy and ICT proficiency are increasing in number and remuneration. Already the hourly wages of women in their 20's are higher than those of men. By comparison, between 1987 and 2002 there was no increase in traditional male jobs and 11,000 fewer jobs in Manufacturing.

Does the Tertiary gender gap matter? It's not an identified concern of the Tertiary Education Commission. There are implications galore if women increasingly become professional and main breadwinner, and men to skilled worker and subsidiary breadwinner - but they are beyond the scope of this paper. What is clear is that if the gap at Level 3 continues to mirror the Tertiary gap, it will continue to grow.

We need a greater understanding of the factors increasing the Tertiary Gender Gap and whether remedial policies are desirable or possible.


MALES

The group most at risk, in this economic transition, is unqualified males. More young males than females have always left school early and unqualified. An Australian study showed push factors to be crucial - poor teachers and the gulf between school and the reality of their lives, revolving around work, cars and girls. Without innovative and largely vocational courses and effective pathways to secure employment such boys may be amongst the males who comprise these depressing and familiar statistics. Males are also more vulnerable to unemployment: in the late 80's unskilled jobs were slashed and the male unemployment rate leapfrogged over the female. These figures are a reminder that the gender gap is about much more than academic achievement. People who express no interest in the gender gap because males still dominate society miss the point; under-educated and under-socialized males also dominate the dark sides of society, to their and everyone else's detriment.

We've now covered items one to four.

Let's look at five.

A very thin antipodean research base suggests that under-achievement – in relation to other males or to females – is most prevalent amongst rural, working class and Maori males.

In preparing for this paper I almost immediately found strong evidence for this – the three pockets of most extreme male under-achievement in New Zealand, both in relation to other males and to girls in those areas, are all rural, low decile, and have a significant or majority Maori population.

I then analyzed each factor in more detail. First, rurality.

In 1999 ERO found that the gender gap was twice as big in small and rural schools as in large urban schools, and four times as large in area schools. ERO concluded that country boys were "disadvantaged". (Of course it's a moot point whether lower performance automatically denotes "disadvantage"). Similar findings have come from Australia: the remoter the district, the great the gender gap, and they should know.

These findings evoke the traditional stereotypes of rural boys aiming low and leaving school early for unskilled local jobs, and rural girls aiming high and leaving for town. I tested these in two ways, anecdotal and statistical.

I spoke to eight rural schools. Their communities varied sufficiently in affluence, ethnicity, labour market and degree of isolation to defy easy generalization. In most, boys did leave earlier. One school had a senior school of 18 girls and 2 boys. There's still more work for men than women in the countryside, and many boys gravitate towards it for lack of visible alternatives. But some principals saw little difference between male and female ambition. Most said the key factor was not gender but parental expectation.

I then calculated the gender gap for schools with under 100 senior students, almost all of them rural. Smaller cohorts naturally produced greater variation but when averaged, the small school gender gap is similar to that of larger schools.

A more thorough investigation of varying types of rural community would undoubtedly find a more complex picture, and perhaps distinguish between the effects of isolation and small size, but, as a generalization, ERO's conclusion that boys are poorly served by rural education, if it applied in 1999, no longer applies.

Gender Gap in Leaving Qualification, by Ethnicity

The gender gaps we have been using have been for all New Zealand students. They in fact camouflage quite different ethnic gaps.

In each category shown, and in almost indicator of educational achievement we could select, the European gender gap is significantly higher than the Maori, while the Pasifika gap varies.

Far from contributing to the New Zealand gender gap, Maori diminish it.

We know, of course, that a smaller gender gap does not mean Maori boys are not of concern. Maori boys are generally the lowest achieving of the eight ethnic/gender groups in New Zealand.

Proportion of students leaving with year 12 or Year 13 qualifications

The Maori gender gap is small simply because Maori girls under-achieve nearly at the level of Maori boys.

We use the gender gap to assist our understanding of overall male achievement because the achievement of females, the other half of the population, with broadly equal intelligence and opportunities, is the most obvious reference point.

But for a minority group, the most valid reference point is the majority. The gender gap is a quite useless tool for understanding Maori male achievement. Maori boys need to be considered in the context of an ethnic gap, not a gender gap.

When members of one race overwhelmingly occupy a certain class it can be difficult to statistically differentiate between the two as factors in any phenomenon. Nevertheless, research suggests that working class male culture also increases the gender gap. Another stereotype is conjured: of boys lost to a macho world of cars, sport and dare-devil or self-handicapping behaviour that often masks low self-esteem and a reluctance to take risks over anything other than speeding. The female working class culture is seen as less pervasive and damaging. For example Ruth Chapman asked in 2000 "why does low socio-economic status seems to affect boys more than girls".

The answer is that, (according to qualification results, at least) it doesn't.

1998 School Cert Decile Gap: In 1998 School Certificate the gender gap is smaller in lower deciles.

2004 NCEA Decile Gap: The same pattern is just discernable for NCEA.

Male pass rate as per cent of female

If we express the data differently, with the male pass rate as a percentage of the female pass rate, again, there's a slight trend for smaller gender gaps in lower deciles.

Possibly race is the complicating factor: the larger gender gap one might expect to find in the lower deciles is absent because most Maori and Pasifika students are in lower decile schools and their gaps are smaller.

For example, surveying lower decile schools, predominantly Polynesian South Auckland has an average gap of 9; predominantly Pakeha Christchurch averages 14.

Class may be a factor at other levels. A Christchurch survey found a larger working class gender gap amongst new entrants, while at University low and medium decile males comprise only 46 per cent of first year students from their deciles, while high decile males comprise 49 per cent of their cohort.

But it can not be concluded that at secondary level class, or race or rurality play either a major or a clear role in the gender gap. Further research might identify a much more complex picture, particularly if it can separate race and class.

We've now covered items 1 to 5. To borrow from Churchill, not the beginning of the end but definitely the end of the beginning...