Education and Poverty

International Educational Data United States Educational Data Texas Educational Data Visualizing Educational Data Michael Marder, Co-Director of UTeach and Professor of Physics
The University of Texas at Austin Introduction and Overview
(Texas Tribune Interviews) Airplanes, failure, poverty,
and the future of US technical
supremacy Availability of data and the Texas SAT College-Readiness Measure Data on Charter Schools A paper on airplane failures and education is at How can the US compete if our schools are not up to par? Other countries educate their citizens K-16
Bring them here for graduate school or temporary work
Hire the best From Science and Engineering Indicators, 2012 By educating our technical workforce for free The opinions expressed on this page are not necessarily shared by the UTeach program, The University of Texas at Austin, the Texas Education Agency, the Texas Higher Education Coordinating Board, the National Science Foundation, or the State of Texas. The importance of foreign-born scientists and engineers to the S&E enterprise in the United States is great.

25% of all college-educated workers in S&E occupations in 2009 were foreign born

42% of doctorate holders in S&E occupations.

The number of temporary work visas issued to high-skilled workers increased steadily until the economic downturn of 2009 and then fell rapidly

155,000 H1B visas issued in 2007, down to 118,000 in 2010

Over 2/3 of H-1B visas were issued for S&E occupations

Asian citizens made up three-quarters of all H-1B visa recipients. http://www.nsf.gov/statistics/seind12/c3/c3h.htm http://bit.ly/MarderAASA Whoa...what's going on here? This is a Prezi. It will look best if you view it in Full Screen mode.
(If you are looking at it in a thin narrow box on the screen, look for the faint gray word MORE on the lower right hand side of the box, click it and select Full Screen).
Hitting arrow keys or the space bar will take you forward on a preset path.
Sometimes it will take you near to images or videos without taking you into them. You can click on them to explore, and then rejoin the path. Educational Data in Three Broad Categories Among 30 developed countries, the United States is ranked twenty-fifth in math and twenty-first in science. When the comparison is restricted to the top 5 percent of students, the United States is ranked last.
-- Davis Guggenheim (2010), Waiting for Superman You might have heard TIMSS--Trends in International Mathematics and Science Study Physics Students, China and US (Bao et al, Science 2009) It's true It's true for college freshmen tested in physics PISA--Programme for International Student Assessment This visualization is based upon a blog posting of Mel Riddile ( http://tinyurl.com/RiddilePP ). He pointed out that the comparison between the US and other countries looks very different if one takes into account the level of child poverty. He also pointed out that PISA disaggregated US schools according to their poverty concentration in four groups, and one could compare these four groups with other countries. Viewed in this way, the US looked as good as or better than any other country.

This was a striking observation and I wanted to check. The weakest point is the calculation of child poverty levels. The data come from UNICEF, and describe child poverty after taxes and transfers. Without taking into account taxes and transfers most developed countries would have very similar levels of child poverty. Once they are included there is huge variability ranging from 2% (Finland) to 23% (US). The definition of child poverty used by UNICEF varies from country to country. The US uses a definition based upon the cost of food. The European Union defines individuals to be poor who live on less than 50% of the median income per person. To justify a standard of poverty that varies from place to place, UNICEF quotes Adam Smith:

Custom...has rendered leather shoes a necessary of life in England. The poorest creditable person of either sex would be ashamed to appear in public without them.… Under necessaries, therefore, I comprehend, not only those things which nature, but those things which the established rules of decency have rendered necessary to the lowest rank of people.

The question of taxes and transfers worried me more. In the most recent UNICEF report on child poverty, data from the US were lacking and the US was omitted. In correspondence with Leonardo Menchini I learned that data had recently become available from the CNEF project at Cornell. Supplementing the Innocenti 2009 report with US data gave me a chance to check how these numbers are calculated.

CNEF is based in turn upon the Panel Study for Income Dynamics (PSID), an income survey that has been running since 1968, and that the US National Science Foundation recently celebrated as one of the most successful projects in its history. It involves the periodic survey of all members in roughly 5000 US families with a focus on questions involving income and finances. In short, the US number on child poverty is not pulled out of a hat, but comes from decades of questioning a carefully composed sample with tens of thousands of members and tabulating the financial resources that flow to them in fact. Whether the data from all other countries are collected with the same care I do not know, but I saw no reason not to use the UNICEF figures with respect.

To complete the chart, one step was still missing. CNEF data told me the fraction of US children in poverty. PISA disaggregated schools according to the fraction of students eligible for free lunch. However students are eligible for free lunch at 130% of the poverty level. Thus I still had to figure out exactly where on the horizontal axis to place for example students in schools where 25-50% of the students were eligible for free lunch (one of the PISA categories).

I solved this problem in two phases. From CNEF data for 2009 I found that 1.4 times as many children were at less than 130% of the poverty line as were below the poverty line. This suggested that I map the interval 25-50% to (25+50)/2/1.4= 27. Now it happens the the PISA scores for the US as a whole are nearly identical to those for the group with 25-50% free lunch, and 23% of US kids, not 27%, are poor. So I decided that my formula still overstated child poverty, and divided by 1.7 not 1.4 to convert the free lunch bins into points on the poverty axis. I believe this estimate bends in the direction of providing a low estimate of child poverty.

Even after all these corrections, the US schools trace out a straight line that lies on top of the other developed countries for which UNICEF provides child poverty data. It is an interesting alternative to the common narrative that the United States ranks nearly last among developed countries in test scores. Notes on
Disaggregation of PISA
Data More confirmation comes from PISA But the story looks somewhat different when countries are arranged by levels of child poverty http://tinyurl.com/PISATBL Examine the data yourself at The story changes again when the range of PISA scores in various countries is displayed according to poverty levels You might have heard that US test scores are flatlining I believe that what is wrong with our schools in this nation is that they have become unionized in the worst possible way....This unionization and lifetime employment of K-12 teachers is off-the-charts crazy
-- Steve Jobs, Feb 16, 2007 National Assessment of Educational Progress (NAEP) Charter Schools Value Added Methods Unfortunately, in state after state, not only are charters worse on average than regular public schools, even the best of them is not truly exceptional. What about charter schools? Don't they provide many families their best chance of escaping failing public schools? But in 8th grade mathematics, scores have steadily risen in almost every state for a decade, and scores have steadily risn for 9- and 13-year-olds since the 1970's And you might have heard that unions are a major part of the problem Data of Kane and Staiger for 78 pairs of teachers in Los Angeles, comparing predictions for student performance on an elementary school mathematics test with actual average student performance following random assignment of teachers to classrooms. After Kane and Staiger (2008), Estimating Teacher Impacts on Student Achievement: An Experimental Evaluation, http://www.nber.org/papers/w14607. Advocates of Value-Added methods for assessing teacher performance commend this as the most careful study. It differs from other studies because pairs of teachers were randomly switched between classrooms before the school year began. Kane and Staiger report scaled scores for which they say that 0.25 corresponds to about a year of learning; this conversion was used to obtain the scale in this figure. They used four years of teacher and student data to form predictions about relative performance of paired teachers in a fifth year. Although measured student performance is correlated with past performance, conclusions about individual teachers were inaccurate.

• Seventeen percent of the time the prediction is too low by half a year of learning ( \pm.125 ) or more

• Thirty-five percent of the time the prediction lies within half a year of learning of the measured value.

• Forty-eight percent of the time the prediction is too large by half a year of learning or more. Comments Video Overview http://bit.ly/MarderNAEP8 Explore the data for yourself at But there is no connection between whether teachers have collective bargaining rights and student test scores There is also a big push to use objective computerized analysis of test scores to tell good teachers from bad Except even the best methods still make mistakes around 1/3 of the time Texas Charters California Charters Florida Charters New Jersey Charters Uncertainties in NAEP scores:
Standard errors for NAEP scaled scores vary from year to year, state to state, and group to group. The standard errors applying to scores of low-income students in 8th grade mathematics range from a high of 4 (New York in 2000) to a low of .8 (Delaware in 2009). Since 2000 most standard errors have been 2 or less; NAEP scaled scores differing by more than 4 are likely to be significantly different. Cost, however, is another matter... http://bit.ly/MarderNAEPF Explore these data yourself at How public schools are scheduled to fail in 2014 All schools are held accountable You might have heard that charter schools produce larger year-to-year student gains Small schools are better, right? Another common claim: the teacher is the biggest influence on student achievement in the school Scan through the graphs below and see for yourself But did you know that Texas has two separate systems with different standards? http://tinyurl.com/MarderTXRescue Explore these data for yourself at On, average, this graph says 'No.' Student Flow Plots Look at how scores after ssI went up and followed higher (blue) paths from 5th to 8th grade. In 8th grade there was a second but less permanent jump. Average Score Change All well-off students who scored 60%-70% on math TAKS in spring 2005. give a way to display tens of millions of student test scores Lesson 1: The performance gap between low-income and well-off students grows fastest at 6th and 9th grades, when they change schools Poverty is no excuse for low educational achievement But there are almost no exceptions to the rule that academic outcomes depend upon poverty concentration, and high poverty schools provide poor odds of academic success Interaction of ethnic group and poverty There is no doubt that poverty presents real challenges, and it is harder to be a principal or teacher in an urban district than it is in the suburbs. But even in the toughest of neighborhoods and circumstances, children excel when the right adults are doing the right things for them.
--Michelle Rhee (2010), Waiting for Superman http://bit.ly/s8XDTX Technical paper at Lesson 2: Texas' Student Success Initiative raised scores for tens of thousands of low-performing low-income students and scores stayed up for years What about graduation rates? http://bitly.com/zJuvju?r=bb Class of 2010, NGA Compact Data from Massachusetts Charters Louisiana Charters Connecticut Charters Pennsylvania Charters New York Charters Focus on New
York's Regents Exams.
Explore the data yourself at http://tinyurl.com/MarderNYRegents Video Maxims Nothing makes sense in education except in light of poverty
When all else fails, assume professionals act in their self interest
Educate every person's child as if it were your own http://bit.ly/VyDH0T A technical paper exploring these matters is at A paper with more commentary on Value-Added Models is at http://bit.ly/SEuuF1 Explore Texas school data for yourself here http://bit.ly/VtWexZ Colorado Charters Illinois Charters