Thursday, January 30, 2014

Educational Attainment in the US, 1940--2013

The last half of the previous century was an astounding time for increasing educational attainment in the United States.  In 1940, for instance, the majority of adults had an 8th grade education or less; 13% of the adult population never made it to fifth grade.  And in 1940, if you had a college degree, you were special indeed: Only 5% of all adults were so privileged. The first interesting point is the effect of compulsory education through high school.

Many people believe the rise in family incomes and quality of life is related to educational attainment, although which is the cause, and which is the effect is subject to interesting debate.  It's pretty clear, however, that the end of World War II was the turning point.

This visualization started out very complex, but I decided to make it simpler in the end: Choose an age category and, if you like, a smaller subset of years.  Compare men to women in educational attainment, and see how that changes over time.

I'm drawn to the 55+ age group: The baby boomers and older.  Notice how college degrees exploded over time for both men and women, but how men in this group are still more likely to have a college degree or more.  Now look at the other age groups: In the younger categories, women have surged ahead, or, depending on your perspective, men have fallen behind.: That's the real story here.  And with enrollment rates of women continuing to be much higher than for men, it's likely to stay that way for a long time.

Tuesday, January 28, 2014

No Graduate Students Need Apply

Often, the data I use on this blog is a fairly comprehensive set of colleges and universities in the United States. But very many very good smaller colleges can get lost in the array of data.

So I thought about colleges that don't award any graduate degrees, and selected the 511 from the list of over 7,000 in IPEDS. These institutions are both public and private, not for profit, and they are in almost every state. Beyond that, though, you get an interesting look at the wide range of diversity in the group via Box and Whisker Charts. If you don't know how to read a box plot, take a quick look at this article, or just know that each dot represents an institution, placed higher or lower on the chart based on the value shown; 50% of the points on the chart are contained within the gray box, 25% are above the top of it, and 25% are below the bottom of it.

Use the filters to include other types of colleges (but be warned, many have not-so-great data) or regions or control.

Monday, January 27, 2014

Colleges and the Weather

A brief aside:

When colleges across the country close for several days in a row, the weather becomes a subject of Higher Ed Data Stories. Using Wikipedia, I drew down data on the all-time high and low temperatures by state. (As an aside, it is really hard to get good data on weather records. Are you listening, NOAO?)

You can choose high temperatures or low temperatures (either Fahrenheit or Celsius; or the range between the state's all time high and low. Although this is a diversion, of course, it's interesting to me because you get surprised when you look at the data, and you find things in a visualization you'd probably not notice in a static table.

Monday, January 20, 2014

The President's Agenda for College Completion and One Little Problem

Many of you know about President Obama's call for the US to lead the world by 2020 in the percentage of citizens with college degrees.  It's a noble challenge to the country: We were first as recently as 1990, but we currently rank 12th.

And of course, it's not just a challenge to students: It's one to colleges and universities to improve access and to improve college completion among enrolled students via a wide ranging series of measures.

I had earlier tweeted that there was one small problem with this: That if we want to improve that rate by 2020, we're going to have to invent a time machine, go back 10 years, and spend more on pre-K and elementary education, which have frequently (but admittedly not definitively, for those who want to argue everything) to be the best investment in education.  Kids who don't do well in school early have greater struggles farther on, and the dream of college is even less likely. (Of course, doing well in early grade school does not ensure college readiness, either.)

And then we have the issue of paying for college, once a student qualifies for admission.  We know that Pell Grants have not kept pace with college costs, and we also know that college costs have risen too fast.  (Note: I include this Pell Grant chart for its information only; the visual suggests just the opposite of reality, in what can only be called one of the worst charts ever.)

One problem: We're pretty much going backwards on something as simple as free and reduced lunches. Take a look at the map below, and pull the slider from 2000 to 2011.  If we have more students from families who can't afford school lunch; rising college costs; and decreased federal support, it's going to be tough to get more kids to college.

Note: The qualifications for reduced for free lunches are not perfectly static, of course.  Here is the data over time.

Monday, January 13, 2014

More on Net Costs

In the continuing debate about college affordability, another way to look at the problem.

This is once again IPEDS data from FY 2011, the most recent available for financial aid and net price.  It shows, on the top chart, the net price for students at every college or university in the US, sliced three ways: Net price for students with family incomes (adjusted gross) over $110,000 in blue; net price for students from family incomes between $48,000 and $75,000 in gold; and net price for students from family income of under $30,000 in red.  They are sorted from the top by the blue bar, in descending order.  For reference, on the right of the top chart is the percentage of freshmen with Pell Grants (in orange).  In congratulating an institution for doing well on low-income students, it's also important to consider how many of them they enroll in the first place.

On the bottom, the same net price data is shown as a percent of family income.  But since we don't really know what the average income is for students in any band, use the blue parameters on the right to estimate. You may think that the average for the Under $30,000 band is $10,000, for instance, or that it's $25,000. It's your choice for all three bands.  The bottom chart updates as you change the input.

Use the filters on the bottom to select smaller groups: You may only want to look at universities in the South, or in Wisconsin, or the most selective, or just research universities.  There's no limit.

Eager to hear your take on this.  Reminder: Income is only available for students who complete a FAFSA; and, as usual, all the caveats about the reliability of the non-adjudicated IPEDS data apply.

Friday, January 3, 2014

Doctoral Degrees Awarded, 2012

I continue to find good and interesting data in the new release of the NCES Digest of Education Statistics. Although it's not yet complete, tables are being published as they are ready.

Again, the good folks at NCES provide a great service, but they still think in terms of 1996: That people want reports they can print off and study.  That may still be true for many people, but I do hope they provide information in a more viz-friendly format at some point in the not-too-distant future.  This visualization is a combination of two of those formatted reports, after considerable scrubbing and cleaning of the data.

It shows doctoral degrees awarded by US Institutions participating in Title IV programs in 2012 by gender, ethnicity, and discipline.  The top chart is a bubble chart (notwithstanding the objections of data visualization experts, for good reasons) and the two bottom charts are bar graphs, colored by ethnicity and gender.

You'll notice right away two things, I bet: First that purple (degrees awarded to Caucasians) dominates, and second, we produce a lot of degrees in law and health.

You can filter by ethnicity or gender or specific discipline if you'd like (all three charts will update), but instead try this: Use the "Degree Popularity" slider to eliminate the top two or three disciplines on the bubble charts.

Lots of good stuff here, if you're wiling to spend a little time working on it.  And yet, there is much that could be visualized that's not.  What would you like to see?