Category: Research connected to practice

Using Data (AMATYC 2017)

For the session “Using Data to Improve Curriculum” (Nov 10), here is the ‘stuff’:

Presentation slides (all):  Using Data for Improve Curriculum

The Handout (shorter): Using Data for Curriculum AMATYC 2017 S116

 Join Dev Math Revival on Facebook:

Transitioning Learners to Calculus in Community Colleges (TLC3)

You might have heard of the MAA project “National Study of College Calculus”  (see http://www.maa.org/programs/faculty-and-departments/curriculum-development-resources/national-studies-college-calculus ).  That work was very broad, as it studied calculus in all 3 settings (high school, community colleges, and universities).

A recent effort is focused on community colleges  with the title “Transitioning Learners to Calculus in Community Colleges”   (info at http://occrl.illinois.edu/tlc3  )  Take a look at their web site!

One component of their research is an extensive survey being completed by administrators of mathematics at associate degree granting public community colleges, including the collection of outcomes data.  A focus is on “under represented minorities” (URM), which relates closely to a number of recent posts here (on equity in college mathematics).

I am expecting that the TLC3 data will show that very few community colleges are successful in getting significant numbers of “URM” students through calculus II (the target of this project).  The ‘outliers’, especially community colleges succeeding with numbers proportional to the local population of URM, will provide us with some ideas about what needs to change.

Further, I think the recent emphasis on ‘pathways’ has actually decreased our effectiveness at getting URM students through calculus; the primary assumption behind this (based on available data) is that minorities tend to come from under-performing K-12 systems which then results in larger portions placed in developmental mathematics.  The focus on pathways and ‘completion’ then results in more URM students being tracked into statistics or quantitative reasoning (QR) pathways — which do not prepare them for the calculus path.  [Note that the basic “New Life” curricular vision does not ‘track’ students; Math Literacy is part of the ‘STEM’ path. See http://www.devmathrevival.net/?page_id=8 ]

Some readers will respond with this thought:

Don’t you realize that the vast majority of students never intend to study calculus?

Of course I understand that; something like 80% of our remedial math students never even intend to take pre-calculus.  Nobody seems to worry about the implication of these trends.

Students are choosing (with encouragement from colleges) programs with lower probabilities of upward mobility.

The most common ‘major’ at my college is “general associates” degree.  Some of these students will transfer in order to work on a bachelor degree; most will not.  Most of the other common majors are health careers (a bit better choice) and a mix of business along with human services.  Upward mobility works when students get the education required for occupations with (1) predicted needs and (2) reasonable income levels.  Take a look at lists of jobs (such as the US News list at http://money.usnews.com/careers/best-jobs/rankings/the-100-best-jobs )  I do not expect 100% of our students to select a program requiring calculus, nor even 50%; I think the current rate (<20%) is artificially low … 30% to 40% would better reflect the occupational needs and opportunities.

Our colleges will not be successful in supporting our communities until URM students select programs for these jobs and then complete the programs (where URM students select and complete at the same rates as ‘majority’ students).  Quite a few of these ‘hot jobs’ require some calculus.  [Though I note that many of these programs are oriented towards the biological sciences, not the engineering that often drives the traditional calculus curriculum.]

I hope the TLC3 project produces some useful results; in other words, I hope that we pay attention to their results and take responsibility for correcting the inequities that may be highlighted.  We need to work with our colleges so that all societal groups select and achieve equally lofty academic goals.

 Join Dev Math Revival on Facebook:

Maybe it’s Not “Men of Color”: Equity in College Mathematics, data part II

A recent post here looked at a summary of pass rates based on “Pell eligibility” and race, where Pell eligibility is used as an indicator of possible poverty.  Take a look at http://www.devmathrevival.net/?p=2791 .  The basic message was that the outcomes for black students was significantly lower and that part of this difference seems related to the impact of poverty.

Today, I wanted to follow that up with some similar data on the role of gender (technically, ‘sex’) in the outcomes of students, accounting for poverty and race.  This seems especially important given the national attention to “men of color” (http://cceal.org/about-cceal/).  As a social justice issue, I agree that this focus on MEN of color is important given the unequal incarceration rates.

However, this is what I see in our data for all Pell eligible students in math courses:

 

 

 

 

 

 

 

 

As for the prior chart, this reflects data over a 6 year period … which means that the ‘n’ values for each group are large (up to 10000 for ‘white’).  Given those sample sizes, almost any difference in proportions is statistically significant.  All three comparisons ‘point’ in the same direction — females have higher outcomes than males, within each racial group.

However, notice that the ‘WOMEN” of color have lower outcomes than men “without color”  (aka ‘white’). A focus on men of color, within mathematics education, is not justified by this data.  Here is what I see …

  • There is a ‘race thing’ … unequal outcomes for blacks and hispanics, compared to white students.
    [This pattern survives any disaggragation by other factors, such as different courses and indicators of preparation.]
  • There is a ‘sex thing’ … unequal outcomes for men, compared to women.
    [This difference is smaller, and does NOT survive some disaggregations.]

There is a large difference in ‘effect size’ for these; the black ‘gap’ in outcomes approaches 20 percentage points (about  2/3 of the white pass rate), while the ‘male’ gap is 5 percentage points or less (90% to 96% of the female pass rate).  In other words, it does not help to be a woman of color; it just hurts less than being a man of color.

I think that pattern fits the social context in the United States.  The trappings of discrimination have been fashioned in to something that looks less disturbing, without addressing the underlying problems.  We have actually retreated in this work, from the period of 40 to 50 years ago; there was a time when college financial aid was deliberately constructed as a tool in this work, and this was effective from the information I have seen.  Current college policies combined with the non-supportive financial aid system results in equity gaps for PEOPLE of color.

Most of us have a small role in this work, but this does not mean the role is unimportant.  If your department and institution are critiquing your impact on people of color, terrific; I hope we have an opportunity to share ideas on solutions.  If your department or institution are not deeply involved in this work, why not?  We have both the professional and moral responsibility to consider the differential impact of our work, including unintended consequences.

 Join Dev Math Revival on Facebook:

Equity in College Mathematics: What does the data tell us about poverty and race?

I am very proud of my department for our decision to do some serious work on equity.  We are having focused discussions at meetings and in hallways, we are bringing up equity in other discussions, and have examined quite a bit of data.  I want to highlight a little bit of that data.  This post will focus on the role of poverty in the pursuit of equity in college mathematics.

Like many colleges, my institution provides access to a centralized data reporting function (“Argos” in our case).  We can use this database to extract and summarize data related to our courses, and the database includes some student characteristics (such as race, ethnicity, and sex … self-reported).  In addition, the database connects to direct institutional records dealing with enrollment status and financial aid.  The primary piece of data from the financial aid record is a field called “Pell Eligible”.

As you know, Pell Grants are based on need; this usually means an annual income of less than $30,000.  Students are not required to apply, even if they would qualify for the maximum award.  However, we do know that students do not receive a Pell ‘award’ unless they have a low income.  For us, this “Pell Eligibility” is the closest thing we have to a poverty indicator.

When we summarize student grades by race and Pell Eligibility (across ALL courses in our department), this is the result.

 

 

 

 

 

 

 

 

 

This graph has two “take aways” for me.  First, poverty is likely associated with lower rates of passing.  Secondly, the impact of race on outcomes is even stronger.  Note that the “Pell” group is lower than the non-Pell group for all races, and that the “Black non-Pell” group has lower outcomes than the non-Pell hispanics or whites.

The situation is actually worse than this chart suggests.  The distribution of ‘poverty’ (as estimated by Pell eligibility) is definitely unequal: 70% of the black group is Pell eligible, while only 40% of the white group is Pell eligible (with hispanics at a middle rate).

I am seeing a strong connection between our goal of promoting equity and the goals of social justice.  As long as significant portions of our population live in poverty, we will not achieve equity in the mathematics classroom … awarding ‘financial aid’ does not cancel out the impacts of poverty.  In addition, as long as some groups in our population are served by under-resourced and struggling schools, we will not achieve equity in the mathematics classroom.  This latter statement refers to the fact that many states have policies like Michigan’s which allow those with resources to have a choice about ‘better schools’, while limiting state funding for public schools (and simultaneously attacking the teaching profession).

In our region, the majority of the black students attending my college came from the urban school district.  This urban school district had a proud history through the 1980s, with outcomes equal to any suburban school in the area.  However, dramatic changes have occurred … even though that district has made significant progress in recent years, there is no doubt that the urban schools are not preparing students for college.  Poverty plays a role within that school district, and the interaction between race and poverty is again unequal: more blacks live in poverty within the city than other races.

The social justice movement seeks to provide all groups with equal access to upward mobility, combined with a reasonably high probability of escaping poverty, based on a presumption of effort.  Barriers to progress are addressed as systemically as possible.  College mathematics is currently one of the barriers to progress in social justice.  Modern curricula do not solve this barrier, given the data I’ve seen (though we are early in that process of change).

If we see our role as separate from equity and social justice, we are enabling the inequities to continue.  This is a set of issues that we can not remain silent about.  Even if we are not committed to social justice, we need to work on these barriers for the good of our profession.  You might begin by discussing social justice issues with your friends or colleagues who teach sociology or anthropology, quite a few of whom have a background in ‘social problems’.

Join Dev Math Revival on Facebook:

WordPress Themes