Guest Opinion | School Board Member Scott Phelps: A Reductionist View

A reductionist view

I remember years ago sitting in the office of our former local state senator who hailed from a wealthy city next to Pasadena.  The senator was asking another board member and I about some PUSD student achievement data and questioning why it wasn’t higher.  Other local and state elected officials have long done this, even publicly, as the virtue-signaling garners them support from voters—-or in the case of advocates, this argument is used to gain more support for certain programs, services or priorities or non-profits that they want funded. (Interestingly, this was the same senator that later didn’t support the funding formula put into place by Governor Brown to give more money to districts with higher percentages of students who were socioeconomically disadvantaged.) I remember another incident, when we attempted to pass a parcel tax years ago.  A public commenter came to object to it because schools in PUSD didn’t all have similar test scores like the South Pasadena schools did.  Recently someone who had just moved to Pasadena wrote to me about PUSD’s overall, aggregate state test scores being too low.  Finally, a public commenter just recently came to us to express concern about our district’s overall, aggregate 3rd grade state test scores in English Language Arts.  The state superintendent recently announced a goal to have all 3rd graders reading at grade level, which is of course a great goal that many have had.  They both noted the correlation between that measurement and later academic success.  

The nature of PUSD

My objection to all of this is the implication that it is valid to talk about any overall measure of PUSD.  PUSD is a collection of very different schools, unlike any of the districts that touch PUSD’s borders, which all have very low percentages of socioeconomically disadvantaged students and schools within them that have similar socioeconomic makeups and test scores.  For example, if one looks at the latest available (2019) English test scores for PUSD’s (at that time) eighteen elementary schools, one sees a very wide variety of test scores, from 24% having met the standard to 80% having done so.  When one makes a scatter plot of this data vs. the data of the percentages of socioeconomically disadvantaged students at these schools, one sees that these scores have a very strong negative correlation with the percentage of socioeconomically disadvantaged students at those schools:

Each point is one elementary school.  The vertical coordinate for that point is the test score, and the horizontal coordinate for that point is the percentage of students who are socioeconomically disadvantaged. The correlation coefficient between these two sets of data is -0.89 which indicates a very strong correlation—the most negative value possible in theory is -1.0. That’s an exceptionally strong negative correlation. If one added to the plot the data from our surrounding districts, the data would fall in the upper left portion of the graph, the portion corresponding to very low percentages of socioeconomically disadvantaged families and higher test scores.  It would fall along the same trend line.  This confirms what we already knew, that socioeconomic status has the biggest effect on standardized test scores, i. e., follow the money. Of course, PUSD educators and administrators have long known about these differences in PUSD schools and their test scores and would say that these data and correlation are no surprise. PUSD’s executive leaders are forced by and state and local politics and by so-called state accountability requirements like data dashboards and the mantra of “continuous improvement,” however, to support the idea that the schools themselves, if they can just measure the right things and improve their services, can overcome these differences and eliminate the so-called achievement gap, which is really just one measurement of a large difference in lives. If they don’t support this dogma, they don’t get hired in the first place, or they don’t get to stay very long in their positions. (In a bizarre state of denial, the state accountability system requires school districts to measure differences in scores between racial and ethnic groups (highly correlated with socioeconomic status and resources) even though the state constitution prohibits them from considering race or ethnicity in public education. And the state senate currently has stated a key value for next year’s budget is to “close [the] learning gap.”)

Just raise the expectations

Those who with good intentions and genuine concern for students’ futures are pointing out PUSD’s relatively low overall, aggregate test scores may know about the socioeconomic status of PUSD families, but they may think that it isn’t as important as hard work and higher expectations and better curriculum and instruction. (And for some, political popularity is a very powerful motivator.)  I am sure by pointing it out I will be accused of having low expectations for those schools that have higher percentages of socioeconomic disadvantaged students and lower test scores.  This was the mantra of No Child Left Behind (NCLB).  We just need to have higher expectations and the problem is the “soft bigotry of low expectations.” That led to some high poverty elementary schools in the region and the nation focusing narrowly on the lower level skills that would get them higher test scores, which they did achieve on those earlier tests. Scripted instruction came into being to help schools meet the demands of the high stakes testing which was narrowly focused on certain skills.  Curricula was narrowed and access to other subjects that didn’t have high stakes tests decreased. That effort has changed into “smarter balanced” assessments in some states such as California, which aim to measure student achievement in a broader way, and in general the tests have become less high stakes across the nation as folks realized that wasn’t a magic cure-all.  There were then and are to this day exceptionally few secondary schools in the nation that have both higher test scores and high levels of poverty. They are mostly charter schools that have high family choice effects, longer hours and more days of instruction, lower pay for staff and very high staff burnout and turnover. The mantra of high expectations, the focus on scripts and narrow testing that worked for some high poverty elementary schools has never been able to broadly overcome the increase in the difficulty of the secondary content and the other facets of the lives of those same students once they reached secondary school.  

Producing a spread in scores and tutoring

The problem is this:  the designers of standardized tests still have to produce a wide enough distribution of scores for their tests to be reliable. The easiest thing to use in test creation that produces such a distribution is academic language that is highly correlated to the student’s socioeconomic status. The tests therefore aren’t purely criterion-referenced, to some sort of objective “standard” as the way tests are reported would have you believe, e. g., “% Meeting Standard.” They are also norm-referenced and the sample used to norm them includes student from wealthy, high-achieving districts.  Studies have shown that students from different socioeconomic statuses enter school with differences in vocabulary in the tens of millions of words. Parents of higher socioeconomic status commonly read to their children from an early age since they have the time and family resources to do that. Such outside of school factors are much more powerful than in-school factors—estimates are that such factors are four to eight times more powerful than in-school factors in affecting student test scores. Most people, including those who sincerely want to help PUSD improve, can’t see these effects because they aren’t in the homes of families of differing socioeconomic statuses.  They are generally familiar with the lives of people like themselves.  Many have access to resources to help their children adapt to higher expectations or more challenging curriculum.  Their corresponding natural belief is that if PUSD raised expectations and improved the level of its curriculum and instruction, students would learn more and would be academically successful.  It’s that simple. That is largely true in the districts surrounding PUSD, but it is also because of outside of school factors.  I do private tutoring for five different agencies.  Nearly all of my students are private school students or students from the districts surrounding PUSD. When these schools have high expectations that students struggle with, their parents hire tutors.  Most PUSD families aren’t in the socioeconomic position to do that.  PUSD schools can’t simply raise their expectations in order to get higher state test scores.  Students would just struggle more, grades would be lower and graduation rates would be lower.  

What could help

What could help is providing more financial resources to lower income families.  This is something that the current federal administration was doing recently, enhancing the child tax credit to lift millions out of poverty.  That enhancement was temporary, and a certain US Democratic senator didn’t believe it was a good idea to increase the child tax credit despite most of his state’s residents needing government assistance.  The bill that he blocked also contained a transformative investment in childcare and early education, which would help with the money issue.  Anything that decreases wealth inequality would likely help decrease the inequality present in test scores.  A study of babies’ brain activity recently showed that just one year of subsidies for their mothers increased the babies’ brain activity in ways associated with stronger cognitive development, see https://www.nytimes.com/2022/01/24/us/politics/child-tax-credit-brain-function.html.  On the related topic of racial inequality, a recent report about such inequality in California, reported here https://calmatters.org/california-divide/2021/12/california-racial-inequality-budget/?fbclid=IwAR0Lea19Tx-wx3jljTM60CJUHHaGZJc6yf8pwRu_shUUbGyVYkZpJLw1-7c, has led the executive director of the California Budget and Policy Center to recommend that more cash assistance, healthcare and childcare services be provided and that it be better targeted to reach communities of color. (Sadly, the corporate California Democrats who control the legislature just killed universal health care without a vote.  Such coverage would have clearly helped racial and wealth inequality and thus most public school children’s educational achievement.)  We will see if the governor and legislature agree to help with this need as they negotiate over the state budget before its June 2022 approval. He does have the state constitution as a barrier, since as noted above it still bans state governmental institutions from considering race or ethnicity in the areas of public employment, public contracting, and public education.

The state of current resource distribution

The state does already provide more funding to districts like PUSD which have a large percentage of socioeconomically disadvantaged families.  And the caring and competent educators and administrators of PUSD have long directed more resources to programs and services to try to address the needs of lower achieving students from lower income families and have tried to implement strategies to try to help raise student achievement. (See the Pasadena Now article at  https://www.pasadenanow.com/main/guest-opinion-scott-phelps-some-facts-about-equity-in-the-pusd

for some examples of greater funding for PUSD schools that have higher percentages of lower achieving students from lower income families.) They do this work to make a difference and have for many years implemented what they thought was best, within state and federal requirements and union agreements, for students to succeed, including interventions for lower-achieving students. In just one example, with regards to reading, the PUSD has for years contracted with a non-profit to bring volunteer readers into the schools to read one on one with students. The Pasadena Educational Foundation (PEF) has for decades provided donated funds and grant-writing to help with these goals and fund many programs and services. In recent years such grants have allowed PUSD to provide health care and other social services at some of the neediest schools under what is called the community schools model. The state in a recent year provided a Low-Performing Student Block grant that PUSD used to fund additional services for families using Pasadena’s Pacific Oaks College as the service provider. In the last few years PEF—through a local donor and his wife, a PUSD board member—and PUSD itself have funded a consultant to work on continuous improvement. There are many, many such examples over the years of PUSD regular and grant spending on programs, including the hiring of many local non-profits annually to provide a variety of services to schools, families and students and also many consultants to try to meet the ever-growing demand for schools to do more and more for students and families.  As folks may know, the federal government has long known that there is a correlation between family income and student achievement and therefore has provided funding for lower income students called Title 1 since the 1960s. Title 1-funded services in the schools have failed to move the needle significantly on test scores of lower income students over the decades, because as stated earlier, in school factors are not nearly as strong as out of school factors. This lack of improvement despite the longtime federal funding was one of the reasons that the accountability movement, e. g., NCLB, came into being.  It should be clear after so many years that raising school spending, spending more on certain school programs or services for certain schools or students, applying data-based accountability or a continuous improvement mantra or any other response that doesn’t change the financial resources of families won’t significantly affect the group differences in standardized test scores that the state measures. It’s the inequality in family wealth that is the greatest factor in producing these scores.  Pasadena is one of the most unequal cities in the state, and of course the nation has a huge amount of inequality.  

A less reductionist approach: Measuring individual student growth

The group differences haven’t changed much over the many years, but many individual students from lower income families have achieved great academic success. Some of us board members have long called for the district to look at those individual students and examine why they were able to achieve highly when the majority of their peers weren’t.  We also feel that following an individual student’s growth is a much more student-centered, affirming and valid way to measure the development of a student’s potential, but the PUSD data system and staff hasn’t been able to do that to date.   Here’s something I wrote on this need for a new way to look at data several years ago: https://thehill.com/blogs/pundits-blog/education/291561-public-school-need-new-approach-to-close-achievement-gap

I still hope that we can change our focus to look at and learn from the progress or lack of progress of individual students, as in the end, individual students are what matters most.  

It’s the inequality

I hope that people refrain from talking about PUSD student achievement as if PUSD is a single entity that can be compared with surrounding districts as that isn’t really valid.  Because of its unique history and Pasadena’s very high level of inequality, it is a collection of very different schools.  I understand that it is easier—and convenient for critics—to lump everyone together into one number, but that doesn’t paint a true picture of PUSD.  I also hope that folks can stop pretending that test score differences aren’t about wealth inequality when that is clearly the dominant factor in test score differences. The accountability movement has for decades pretended that schools themselves can erase these differences. That just isn’t true.