Wisconsin Policy Research Institute Report:
By Sammis White, Ph.D.
by Gender, Race, Grade, and Income
Given the smaller number of students in the categories of “reduced or denied” lunch status and their common results being somewhat in between free lunch and no support, we reduce the basic comparison to just those with either free lunch or no support. This will make differences more pronounced and keep distinctions clear between low-income students and middle-income students. Again, we look at reading and math separately.
First, we look at reading, starting with fourth grade to see if there are large gaps by gender, even among those of similar incomes. We start with the fourth grade and examine average reading scores by gender, income, and race. There are six entries on the first line in Table 14. This line contains the average scores of students who qualify for free lunches, meaning they are lower income. What is immediately obvious is the pattern discussed above: in reading, males are behind comparable females in average reading scores regardless of their income. The largest gap among the lowest-income students is between white males and females, followed by African-Americans. The key point for this report is that once again, African-American male scores are both below African-American females’ scores and below all other scores.
When we examine the scores of middle-income students (no support), gaps between males and females exist regardless of race. On reading, males, on average, do not do as well regardless of race or income at fourth grade. When we look across grades, we see the same pattern: males are consistently below their female counterparts on reading across all grades (Table 15). In most grades, the male-female gap is pretty similar. There are some anomalies, but male averages are basically well below female averages regardless of race or income.
A further point to note about the African-American males is that the low-income males drop further and further behind the middle-income males (Table 16). The gap between the two starts at -22 points in fourth grade and slowly moves up to -38 points by sixth grade, drops back a bit and ends at -34 points in tenth grade. That it a substantial difference across income groups. But the gap is even wider among white males at tenth grade: 48 points separate low-income from middle-income. But to put that in perspective, the low-income, African-American male reading score, on average, is an additional 38 points below low-income, white male’s average 10th grade score. In other words, by tenth grade, the average low-income African-American male scored 86 points below the average white male with incomes too high for lunch support. That is several years of learning and clearly points to a problem.
We look next at math and use fourth-grade average scores as an illustration of what is found. Math scores across three races vary by income (Table 17). But the scores differ little by gender, except among African-American students. Once again there is a gender difference. Low-income, African-American females in fourth grade outscore low-income African-American males, on average, by three points. As with reading, scores by gender and race rise with income. Across races, the lowest scores are achieved by those on free lunch, followed by those with reduced-price lunch and those denied a subsidy, regardless of race.
Table 18 shows the gender gap in average math scores by race for low- and middle-income students by race. The differences in average test scores by gender do not seem to be race-based. By examining those with sufficient incomes to not be in the lunch program, we can see that among all three racial groups there are very modest differences in average scores between genders, with an occasional exception. One exception is among sixth-grade Hispanic students; the second is tenth-grade African-American students. These have 16 and 14 point differences, respectively. Such a difference is odd, since many other years have 2-4 point differences. We attribute this to a small number of students in each of these cells. The basic pattern is that males and females score relatively similarly. Thus, for the most part we must conclude that although middle-income, African-American males score consistently below African-American females with similar incomes in math, the differences are not pronounced, on average, until tenth grade. The issue is not an early problem across gender, although African-Americans are the only racial group in which males are consistently behind females.
The Class of 2011
Statements on levels of achievement should not be made on the basis on just one class, even though the number of students in that class for which we had complete information was close to 7,000. To determine whether the same pattern holds true and to see if African-American males are behind African-American females and others earlier than fourth grade, we chose to examine another MPS class, the class that is scheduled to graduate in June, 2011.
Because this class has not progressed as far in school as the Class of 2008, test data are available only up through seventh grade. The advantage is that the scores start at second grade. This is the only class MPS tested at second grade. Unfortunately, there is a hole in the reading scores because of the use of a different reading test in third grade that is not calibrated in the same fashion as the other scores. That said there are still lessons to be learned from examining the data from another class.
The key point to note in Table 19 is that on reading, males consistently have a lower average score than females across all of the grades. The difference starts at seven points in second grade and grows to 13 points in seventh grade. This pattern and scale of difference are exactly the finding for the Class of 2008.
Table 20 reveals the pattern of average math scores for all students in grades second through seventh As the reader may recall, among students in the class of 2008, the girls, on average, outscored the boys on math in grades 4th through 10th. But the differences never exceeded six points. Little can be said of the differences: the two genders’ scores were almost the same. In the class of 2011 male
math scores were higher than females in fourth and fifth grades and lower in sixth and seventh. That is hardly a strong pattern, especially since the difference never exceeds three points. So both classes are similar in math: difference in gender is not related to average score.
As with the Class of 2008, race does seem to be associated with different levels of reading achievement (Table 21). Average white scores are highest, followed by Hispanic and then African-American. The gap between African-American and Hispanic is present but relatively small by gender in the second grade. But the differences grow rather dramatically between second and seventh grade by which time African-American females, on average, are 19 points behind Hispanic females and African-American males are 28 points behind Hispanic males.
Gender differences in the two MPS classes also are similar (Table 22). African-American males in every grade score, on average, lower than African-American females. For the Class of 2011 the gap is nine points at second grade, rises to ten points by fifth grade and is 16 points in seventh grade. The gap widens; it does not narrow. A similar pattern holds for the Class of 2008.
The same pattern of increasing gender differences holds for whites and for Hispanics in both classes. For Hispanics in the Class of 2011 the gender gap is not as wide, but it is certainly present in terms of reading achievement. It is possible that the males made up a little ground on their counterparts in the Class of 2008. Among whites the growing gaps are very similar in the two classes. Overall, the gender gaps’ presence and scales for the Class of 2011 are very similar to those found in the Class of 2008: males start behind and drop further behind females as they go through school. Unfortunately, this is most pronounced among African-Americans.
Average math scores by race and gender for the Class of 2011 are quite similar in many ways to those of the Class of 2008 (Tables 23 and 24). For example, African-American scores are lower than both Hispanic and white, and year-to-year gains in average score by race and gender are often about the same size. But unlike the Class of 2008, African-American males do not always have lower scores than their female counterparts. The gaps are not large, as we see below (Table 24).
In math the results for the Class of 2011 are a bit different from those for the Class of 2008 (Table 24). Rather than having gaps as high as eight points (sixth and seventh grades, Class of 2008), the gaps in 2011 never exceed five points. And instead of males always being lower, average male scores exceed female scores in third and fourth grades (two points and one point, respectively) in the Class of 2011. The fourth grade net difference in math gap between the two classes is five points. While appearing to be dramatic in terms of males exceeding females, the scale of the difference is sufficiently small as to not be very meaningful. Thus, on math, across grades and races, the two classes are relatively similar in their findings among African-Americans. Males, on average, do lag, and the size of the gap increases over time after fourth grade.
Among whites and Hispanics the basic assessment that genders are quite similar still holds. But there are differences across the years. In the Class of 2011 males exceed females, on average, across grades second through seventh in math. The same pattern holds for Hispanics. This was not true for 2008. But again the differences in scores are often so modest that they are negligible. (There are a couple of exceptions, but they may well be due to smaller numbers of students.) The basic point is that over most grades there is not a real gender differential in math between males and females within these two racial groups. Among African-Americans the gender gaps grow and become more substantial by the seventh grade. But the real differences between genders by race appear in reading.
Income and Gender Combined for African-Americans
The next question is whether the pattern seen for the Class of 2008, low-income African-American males scoring somewhat below African-American females on math and considerably below African-American females on reading are similar for the Class of 2011. We will only talk of grades 4-7, because those are the only grades for which we have comparable scores.
Having shown that there are achievement gaps between genders among the three races, especially on reading, we turn next to focus on the differences in achievement by race within genders and across incomes. The point is to explore just how different African-Americans are from whites and to see if these differences are similar across the two income extremes, low- versus middle-income in the MPS student population. We use the Class of 2011 to explore the differences.
The easiest way to examine racial differences within genders is to focus directly on the gaps. Thus, Tables 25 and 26 show the average scores and gaps by grade of African-American and white females and males. The gap is the gender gap by race within the same income category.
African-American females may, on average, be ahead of their male colleagues in reading, but they are substantially behind white females, on average, in every grade, fourth through seventh. African-American males are also, on average, markedly behind white males in every grade. Income matters little in most grades: the scale of the difference between those eligible for free lunch and those who are middle income is negligible in almost every instance. The one pronounced difference is seen in seventh-grade males, where African-American, middle-income males have fallen even further behind white male counterparts. Notable are the size of the Black-white gap and the fact that the size of the gap continues to rise as grade levels rise. African-Americans are behind by the fourth grade and the gaps in reading achievement grow as they continue in school.
Another important point to note is that the gaps between African-American and white females and African-American and white males are very similar in size at each grade level. That reinforces the point that the real issue here is not one of gender. Minority males and females are equally behind their white counterparts, on average.
What also must be examined is the gap between the average scores of the highest achieving group, middle-income, white females and the lowest-achieving group, low-income, African-American males, to learn the scale differences that exist (Figure 2). For this class in fourth grade, the gap was 49 points (662-613). At sixth grade the gap between average middle-income, white females and low-income, African-American males was still 49 points (672-623). This gap increases to 70 points (525-455) in seventh grade (using scores from tests that were calibrated differently). In short, there are large score differences between the averages of these two groups that do not narrow over time. Middle-income, white females have a huge lead over low-income African-American males in reading, and the gap likely increases even further after grade seven, if the Class of 2011 is like the Class of 2008. In the early years the gaps seen in the two classes are very similar.
When we make African-American versus white comparisons across income categories and within genders for math, we see gaps in averages that are as large or larger than was found in reading (Table 26). At fourth grade, low-income, African-American females are 17 points, on average, below low-income, white females in math, and low-income, African-American males are 20 points below low-income, white males. Those differences grow as the students move up through the grades. What start out among low-income students as 17 and 20 point differences in fourth grade averages grow to 30 (female) and 36 (male) points by seventh grade. Among middle-income students the gaps are very similar in seventh grade (31 and 38 points, respectively). But the gap is not quite as large in fourth grade among the females (13). Just as in reading, there are large differences within genders in achievement in math between African-American and white students in MPS in grades 4th through 7th.
The math gap is substantially larger when we again compare low-income African-Americans males with middle-income white females (Figure 3). At fourth grade 30 points (634-604) separates the two, on average. That is saying that the gap is at least one year by the fall of fourth grade. By sixth grade the difference between the two groups is 44 points (666-622). And by seventh grade the difference is 61 points. Low-income, African-American males are, on average, well behind their middle-income, white female counterparts before they even get to high school. The scale of that difference goes some way in explaining different high-school graduation rates. Race, income, and gender all contribute to these substantial gaps.
The basic conclusion from these many comparisons of averages is that the basic patterns of African-American males scoring below African-American females and everyone else is very true on reading, regardless of the grade level. Even at second grade the difference in averages is ten points. On math, however, African-American males cannot be said to be behind their female counterparts. On the other hand, both, on average, are markedly behind not only white students but also Hispanic students. Those differences grow as the students move though the grades. So, even if the male and female African-American math students do not differ much in achievement through several of the early grades, they fall increasingly behind other students over time. This same pattern is found in the Class of 2008.
The many findings across two MPS classes suggest that the issue in reading is early struggles and low levels of achievement, especially among males. The distance males fall behind females, on average, appears very hard to make up, regardless of race or income. But of greater import is that the gap in averages across incomes, regardless of gender is even harder to reduce. In fact, the gap widens substantially as children age within each racial group. Thus, important as gender may be, it only exaggerates the much larger differences in achievement associated with income variation. Race does play a role, as there are differences in achievement within the same income level. This finding would suggest a need for interventions are not aimed just at one gender or one racial group.
The findings developed above on differential levels of student achievement in MPS make three key points. The first is the score differentials make an extremely strong case for taking action, action that is far more focused and dramatic than has been taken to date to address these many achievement gaps. The gaps between genders and among races and income levels are unacceptably large. Second, the data also make a very strong case for special attention to be paid to African-American males, especially low-income, African-American males who constitute between 25% and 30% of the MPS student population. This group consistently underperforms others in the district. Third, the data also make an even stronger and inclusive case for taking the necessary steps that help all minorities and actually all children in MPS achieve at higher levels.
On the first point, there is a sizable gap in reading achievement for males compared to females and a more pronounced gap between African-American males and females. That said there are larger gaps between African-Americans and whites, regardless of gender. When compounded by gender, the gaps are extremely large. The gaps expand even further when income is taken into consideration. Thus, low-income, African American males are, on average, years behind middle-income white females by seventh grade and even further behind by tenth grade, be it reading or math.
The scores examined in this report show that African-American males, on average, start their educations behind even African-American females. The males never catch up, on average, in reading and seldom catch up in math. The males are years behind middle-income white males and even further behind white females in MPS. If measured against the average student in the state of Wisconsin at 8th grade, the average African-American male in MPS is approximately 60 points behind in math and 49 points behind in reading.[i]25 Those are huge gaps that should be totally unacceptable to the citizens of the state.
On the second point, there are large ethnic differences when gender and income are compared. Though very large for males, the gap is almost as large for females. African-American females, though doing better than African-American males, are not doing that much better. Low-income, African-American females are significantly behind (55 points) middle-income white females, on average, in seventh grade (Class of 2011). MPS African-American females, with an estimated high school graduation rate of 46%, suffer at half the state’s high school graduation rate. The size of these gaps strongly suggests that although male needs for attention are great, African-American females also need a good deal of attention. Both gaps are far too great to leave these groups in schools that are following the same basic approaches that have kept these gaps in place over time. Dramatic steps must be taken.
And third, when we see Hispanic male and female graduation rates just above those for African-Americans and average student test scores above African-Americans but far below comparable whites, a strong argument can be made that both minority groups and both genders should be the focus of all efforts at raising student achievement levels. The key is to focus attention on these gaps and take explicit actions to raise both test scores and graduation rates of all minorities.
Furthermore, the gaps between low-income whites of either gender and their middle-income white counterparts suggest that low-income whites should be included in the target groups for new initiatives. The achievement gap and the gap between white graduation rates in MPS and white graduation rates in the rest of Wisconsin show that there is a long way to go before they become equal.
The average score differentials shown in the analysis above clearly illustrate the need for more dramatic change in how children in Milwaukee are educated, both before and during K-12 schooling. With the knowledge that low-income children start formal schooling behind and fall further behind middle-class students and with the further knowledge that low-income, African-American males start furthest behind, it is critical that new steps be taken to reduce the achievement gaps. Exactly how to reduce and eventually eliminate these gaps is not fully understood. But there is evidence that several approaches can make a difference, if they can be implemented.
This report cannot review all of the options; it just makes the case that African-American males are behind all others and that the gaps are shameful. As shameful are the gaps between white and most minority educational outcomes. What should also be unacceptable are the gaps between low- and middle-income whites. Dramatic steps must be taken to shrink all of these gaps. Here are a few ideas to be implemented before and/or during K-12.
MPS has struggled for many years to bring up low levels of student achievement. But standardized test scores reveal year after year of lower percentages of students at or above proficiency when compared to the rest of Wisconsin. A good deal has been written about this, and Choice schools and Charter schools gained support because of the gaps between desires for higher levels of student achievement and reality. Efforts have been made within MPS to change. But the pressure to really succeed has not been sufficient to bring the degree of change and commitment that is necessary for success.
Educational outcomes in MPS have largely been ignored because of labor surpluses. Employers have complained about the shortage of capable workers coming out of MPS, but they have not really felt much pain, as they have been able to coax more women into the workforce (Wisconsin has the highest female labor force participation rate in the nation) and have benefited from the application of technology. Employers have also benefited from modest levels of immigration. But the world is changing. The U.S. will soon face a huge worker shortage, estimated to be 10 million by as soon as 2010. Wisconsin will experience at least a proportionate amount of the pain.
Milwaukee is already hearing complaints from employers that they have worker shortages. Some of these open positions are for those with few skills. Many positions could be filled by those with just high school degrees. But because of decades of non-graduations, Milwaukee has a very modest number of those with degrees who have not been able to find work. The region needs more high school graduates. Milwaukee is home to over 40% of K-12 age individuals in the metropolitan area. But when approximately half fail to graduate, that means that more than one-fifth of the possible metropolitan workforce is not available, exacerbating the worker shortage.
Milwaukee cannot afford to wait around for very incremental change, which is the best we have seen from overall test scores over several years. Milwaukee must take steps that will make more of an impact on outcomes. Obviously, that will not be easy, as the district has struggled for at least 20 years to match its previous successes.
Given the very compelling case for dramatic intervention, the question then is what can be done to reduce and eventually eliminate those gaps by income, race, and gender. That question and many variations on it have been asked for decades, with only occasionally accurate answers. A few schools have done well with this urban, largely low-income, African-American male population. Some of the successful schools exist in the city of Milwaukee. Their success proves that success in this setting is possible. The lack of replication of their success proves that copying that success in other schools is extremely difficult, especially with low-income students. We have more than ample proof of that statement.
There are several steps that are very likely to make success easier to achieve in K-12 education. Some of these steps are harder to achieve than the others, although the reader may have trouble deciding which of the five mentioned is really harder—since all will be opposed to at least some degree. The five recommendations are:
Increase Parental Involvement in Their Child’s Education
Parents are a critical component to success for which some interventions can successfully substitute: we have examples of individuals who have succeeded without much parental support and whole schools that have succeeded without much parental support. But these have proven to be exceptions. It is easier, usually much easier, to achieve higher levels of learning with parental support of education and the work required.
Greater accountability for MPS outcomes is needed at many levels. And it must start at the top. The state must step up and take greater responsibility for the outcomes. The state invests $900 million a year in MPS and asks for nothing. That is irresponsible. The governor and the legislature must be responsible for the outcomes and take steps to ensure that better outcomes are achieved. The future of the state depends on this. That truth is stated repeatedly in both Milwaukee and Madison, but little is done. It is time for real action.
There are a number of steps that can be taken in the community, at the schools, and at higher governmental levels to begin to immediately address the young, African-American male achievement gaps by forcing greater accountability on those who should be responsible for the outcomes. These include such steps as:
Replicate Successful Schools
Other, more specific steps can and should be taken at the schools to contribute to greater student success. The first of these is replicating successful schools on a much larger scale. We have examples of schools in Milwaukee that have succeeded, for example, Barton, Bethune Academy, Clarke, Hawthorne, Maryland Ave, Meir, and Milwaukee College Prep, to name a few. Others are having success adding value, even if the results leave students behind where they should be. But there is still insufficient duplication of the best schools’ successes. Replicating them goes well beyond just duplicating a few elements. That is a tall order, given the many elements pushing against replication. But replicating these schools and their success should be a clear goal. And greater than current efforts are required.
In fact, it will take an extremely concerted effort, the support for which has yet to appear. Some of the essential elements have been identified: well-qualified and well-prepared principals who are extraordinary at leading; well-qualified, well-informed, and extremely committed teachers who work with the belief that all students can succeed; sufficient resources to offer the tools, courses, and persons needed for success; often the use of particular curricula that have proven success when implemented well; and support for the individual student in terms of nutrition and health care, at a minimum.
Better Prepare Children for School
Since African-American males are behind by the time they are tested in second grade, it seems very logical that efforts be made to help them come to school as ready as anyone else. This implies that they should be attending preschool. It should not be just any preschool but one that actually works with them to assure that they are ready to excel once they arrive in K-12. These children should be enrolled in top quality early childhood education, either in MPS or outside. Furthermore, they should be given access to other services, such as meals and health care that are likely to speed their development.
Embed Body Movement
As part of student preparation an intermediate step that is beginning to build a research track record is the initiation of body movement exercises for students, especially in elementary and pre-elementary classrooms. Body movement is not traditional physical education. It is scripted and focused use of body and limbs that is explicitly designed to build synapses in the brain, so that children have in place the neuron connectors that allow them to succeed at academic subjects. This is not time consuming or aimed at only a few students, although it can be. It is intended for all students for a period of minutes a day.
Many children today, especially inner-city children whose outdoor play space is limited and whose outdoor time is often limited because of fear of violence in their neighborhoods, do not do the activities that develop the ability to read across the page, for example, much less comprehend what they read. Childhood activities today often concentrate on one side of the body and hence, one side of the brain. Using one’s thumbs for a video game does not develop one’s ability to cross one’s arms or legs or make one’s eyes read across the page.
Research done elsewhere makes a good case, but data are also available locally. The recent application of these movement exercises in five, third-grade classrooms in an inner-city Milwaukee elementary school yielded promising results. Upon entry in the fall of 2006, 41% of these students were reading at or above grade, as judged by a simple test that teachers use to understand generally where student skills lie. By January 2007, after one semester of the usual curriculum plus the body movement, 61% of these students were reading at or above grade. That is an increase of 49%. Comparable student scores from another school are not yet available for comparison, but on face value that is a large increase in achievement in just one semester.
And it was not just students
at the upper level that made sizable gains: there was a 61% drop in the
percentage of students reading at the lowest level
(pre-primary/primary/emerging) and a 25% decline in the number of students
reading at the first- or second-grade levels. Those are dramatic
improvements that strongly suggest that this program of movements should
be part of many more children’s daily activities.
The MPS schools are extremely challenged by many discrepancies in student achievement. Gender gaps are one of these challenges. But even greater challenges are racial and income differences. These differences have been identified for years, yet not enough has been done to address them. Milwaukee can look to other cities and think that Milwaukee is no worse off than Chicago, or Baltimore, or Newark, or Detroit. But that does not solve the problem. There are huge gaps in student learning that must be addressed and addressed soon. These students’ futures depend on it. The Milwaukee economy depends on it.
The options briefly discussed above deserve some attention. More must be done now to address these many unnecessary differences in student achievement. The gaps are much too large to ignore any further. Milwaukee and Wisconsin must step up to jointly eliminate these enormous differences. Both entities have too much at stake to allow these conditions to continue as they have.
[i] In the fall of 2002 the average scale score for the State of Wisconsin in eighth grade was 686.9 in reading and 704.9 in math. Low-income African-American males in MPS had respective average scores of 638 on reading and 645 on math. Source for Wisconsin scores: http://www.dpi.state.wi.us/oea/xls/kcmnsclscr.xls.
[ii] These recommendations are taken from a report by Renku Sen, Executive Director of the Applied Research Foundation: (2006). A Positive Future for Black Boys: Building the Movement. Cambridge, MA: The Schott Foundation for Public Education.
[iii] A recommendation for high-quality, very early-childhood education, especially for low-income children, has a great deal of research support behind it but very limited political support to date. If this quality early childhood education was to be available, it should be available to males and females, minority and majority. The research shows that such efforts, if done with quality instruction and support, can yield benefits of $4 to $16 per each dollar invested in the program.
This should not be modeled after Headstart, a program that has an unimpressive track record. It must be more comprehensive, taught by more highly educated staff, and held to higher standards. It will also be more expensive. There is growing knowledge of this subject. And there is Wisconsin support slowly building through such efforts as http://www.wisbiz4kids.com/, a business organization trying to build political support for investing heavily in early childhood education in Wisconsin. Such organizations need support. That apparently will take time. But the evidence on the contributions of early childhood education to higher achievement levels, higher high-school graduation rates, less involvement in crime, and better employment outcomes make a compelling case that this initiative should be pursued.
[iv] Several studies suggest that learning complex movements stimulates the part of the brain used in problem solving and learning. One example that employs a series of cross-lateral movements (right elbow to left knee, etc.) is a fundamental program named Brain Gym. It employs a 26-movement series for use in interested schools and companies. Time spent on this type of exercise can be as little as 15 minutes/day and can be done right in class.
Independent research indicates that participation in Brain Gym can be linked to higher test scores, less hyperactivity, better concentration, memory improvements, and better relations between student and teacher. Post-tests in one school showed a one-to-two-year growth for all students on the reading and comprehension testing and growth of one or more years for over 50 percent of the students on math scores—greater results than might have been expected for Special Education students. Behavior patterns also improved.
©2007 Wisconsin Policy Research Institute, Inc. P.O. Box 487 Thiensville, WI 53092