In November 2023, the U.S. Department of Education released new data about bachelor’s degree completion rates for transfer students. I have been interested in transfer students for years and it is a topic that I want to better understand. I use this blog to remember the data I read about and capture my initial thoughts. It is also a place for me to explore using data visualization tools.

DYAD

The researchers defined a dyad as a pair of institutions consisting of a public community college and a public or private four-year institution. The data only includes pairs where at least 30 students enrolled in the community college in 2014 and at least 30 students transferred and graduated from the four-year institution in at least 8 years. Nationally they identified 385 dyads, of which 8 were in Maryland.

Success of DYADS

Montgomery College belongs to four of the eight dyads in Maryland, showing that Montgomery College has a robust transfer program.

Maryland’s most successful dyad is students who transferred from Wor-Wic Community College to Salisbury University; 10% of students who transferred using that pathway graduated with a bachelor’s degree. Nationally, there were only 18 dyads with completion rates of more than 10%. The most successful dyad, Tri-County Technical College X Clemson University, had a completion rate of 20%. Kapiolani Community College X University of Hawaii at Manoa had a completion rate of 16%. Followed by four dyads with a completion rate of 13%, four dyads with a completion rate of 12%, and eight dyads with a completion rate of 11%.

Size of Transfer Programs

In addition to the completion rate of students transferring in a dyad, it is interesting to examine the size of a dyad program because in my mind a truly successful dyad will have both a high completion rate and be the "right" size. What the "right" size is I'm not sure of yet, but of a size that shows that is sufficient to support the continued transfer of students and meets the needs of the students.

The largest number of students transferred from the Community College of Baltimore County to Towson University (2,282). It looks like all transfer students from Montgomery College who transferred to a four-year institution are counted in all four Montgomery College dyads, since the number of transfer students in the denominator is 1,856 for all four dyads. This is not something I fully understood when first looking at the data. If this is true, Montgomery College has a very high total completion rate of nearly 17%.

I experimented to make a chart showing both the number of students who transferred to a four-year institution and the number of students who completed a bachelor's degree. I sorted it by percentage of students who completed a bachelor's degree. I believe the chart works fairly well at illustrating all the data, and it helped me see the potential issue with the Montgomery College data at the same time.

Immediate college enrollment decreased by 5 percentage points for both low-income and non-low-income HS graduates of the class of 2022.

Dual Enrollment College Pipeline

Twenty-one percent of public school 12th graders in Maryland during the 2019-2020 school year had a college record, that is they had been enrolled in a college class. The majority of these students participated in a dual enrollment program through their local school system. In fall 2020, the fall after they were slated to graduate from high school 79% of students who had a college record enrolled in college. This was during the height of the COVID-19 pandemic. Of those who enrolled in college, 75% enrolled at an in-State institution and 25% enrolled at an out-of-state institution. A little more than half of the in-State students (55%) enrolled at a community college, the remaining 45% enrolled at a public four-year or State-aided independent institution. Further, approximately half (51%) of students who enrolled in an in-State institution enrolled at their college of dual enrollment.

Source: Maryland Longitudinal Data System Center

Further Questions

I would be curious to know how college enrollment patterns of dual enrollment students compare to all public school 12th graders in Maryland. The Sankey graph above with that information would be much improved. Without that information, I am not really sure what it all means.

I am also interested in whether the dual enrollment credits earned transferred to the college in a meaningful way. According to MLDS data, students that took dual enrollment courses in high school earned an average of 2.14 credits.

Further, I wish that I had information on whether these students took AP or IB courses, which can also lead to college credit. I am curious if students are taking dual enrollment courses in addition to or instead of these courses. I wonder if the courses taken are primarily due to student choice, or due to the courses available at the student’s particular high school.

Finally, I am curious how these students perform in college.

This drill-down sunburst chart shows the breakdown of students that transferred to the University of Maryland College Park in fiscal 2020. The first layer shows the type of sending institution: Maryland Community Colleges; other University System of Maryland institutions; Maryland private institutions, including institutions that belong to the Maryland Univerity and College Association (MICUA); and out-of-state institutions. The second layer shows the institutions. However, since the original data source only shows the top institutions in a particular category, the names of not all out-of-state institutions are captured.

This chart is based on an Amcharts demo for a drill-down sunburst chart. This particular data only had two levels, but additional levels can be added if you have hierarchical data with multiple levels. As with several other charts I have tried making with Amcharts, this chart did not work in their add-in for WordPress. I however was able to add my data using Notepad. Then I saved the document as an HTML file by changing the file extension to .HTML. Then I uploaded the file to the media library for my website. Finally, I used some HTML code on this page to creat a frame and pull up the chart. I explained that process and shared the code I used in a previous post about my experiences making charts and maps with amcharts.

The chart shows that 63.6% of transfer students came from Maryland community colleges. Half (50.3%) of those students from Maryland community colleges came from Montgomery College, which is located in Montgomery County not far from UMCP. Anne Arundel Community College, Prince George’s Community College, and Howard Community College were also the source of a large number of transfer students. The second-largest source of transfer students is out-of-state institutions. For that category none of the sending schools are dominant. The biggest category is “other”.

UMCP Transfer Students in Fiscal 2020 Drill Down Sunburst Charth

Source: University System of Maryland, IRIS

These charts shows the student level: freshman, sophomore, junior, senior, or unclassified of students that transfer from Maryland community colleges to University System of Maryland institutions. According to the notes on the data by the University System of Maryland, the transfers are the number of undergraduates enrolled for the first time at the institution with known prior undergraduate post-secondary experience. The students may or may not have transferred credit.

Treemap Maryland Community College Transfer Students by Student Level

Source: University System of Maryland, IRIS, Maryland Community College Transfer Students by Student Level, Fiscal 2020

Drill Down SunBurst Maryland Community College Transfer Students by Student Level

Source: University System of Maryland, IRIS, Maryland Community College Transfer Students by Student Level, Fiscal 2020

I found code for a drill-down sunburst chart on Amchart’s website. I just popped in the data I had already typed up that has Bowie State Univesity spring 2021 headcount enrollment by area of origin. This version is kind of nice because you can drill down to see the details of the smaller categories. A disadvantage is that you can not see the drill-down categories until you drill down. In the future, I may be able to figure out how to show the hidden layers, but I am not sure if that will be an advantage, it might be too close together to see the categories well.

To drill down to a level click on the pie slice. This will expand the slice to the entire area of the pie. It is rather difficult to explain, it is easier to just play with the chart down below to understand how it operates.

The drill-down sunburst chart goes down to the level provided in the original data. For in-State students, the data shows to the county level the area of origin. For out-of-state students, it shows students from the major surrounding states and the District of Columbia. The states are New Jersey, New York, Pennsylvania, and Virginia. The remaining students are from other states, “foreign”, or of unknown origin.

To make the other sunburst chart look cleaner I divided the counties into five regions: Capital Region, Central Maryland, Southern Maryland, Western Maryland, and Eastern Shore. The regions and the counties assigned to each region are from the “Visit Maryland” website. For this semester Bowie State University had no students from Western Maryland counties, so that region is not on the chart. As with the other charts, I withhold judgment on the chart until I have an opportunity to play with it.

Bowie State University Drill Down SunBurst CHart

Source: University System of Maryland, IRIS, headcount by area of origin

Since I was able to get the sunburst chart to finally work properly. I decided to take a look at the data I entered. For some reason, the enrollment for winter 2021 is really low (COVID and short term?), so I decided to pull numbers of a more typical term (still COVID) to get a better idea of actual enrollment. I wrote the code in a text file, so I have not yet taken a look at the chart. However, I know from just looking at the data that Prince George’s County has the highest headcount enrollment. This is logical because Bowie State University is located in Prince George’s County.

Bowie State University Spring 2021 HeadCount Enrollment

Source: University System of Maryland, IRIS, headcount enrollment by area of origin

Getting the COde to Render Locally

In my last post, I said that I could not get the Sunburst Chart to render properly out of CodePen. I walked away with plans to go to bed. Getting ready for bed I did a few minutes of googling and found a Reddit post from three years ago that explained my issue. Apparently, since it was running locally I need to add “https:” prior to the resources in the HTML part of the code.

Below I have quoted partly how it was explained on the Reddit post. Basically, without the https:, which is sometimes assumed and can be assumed when it is running off CodePen, the computer does not know what it looking at. Thus, my code will not render. Since I am new to this and kind of just trying to make things work, I will be adding in the “https:”

Quabourter

…when you don’t use an explicit scheme, then your brower will use the scheme of your page. … This is also why this gave you problems when running locally: if your page is served under file:///some/path/to/your/file.htlm then the url// mascdn.bootstraphcdn.com/font-awesome/4/2.0/css/font-awesome.min.css resolves to file://maxcdn.bootstrapcdn.com/fount-awesome/4.2.0/css/font-awesome.min.css, which doesn’t exist. …

When I was running the code without the “https:”, it would just spin its wheels. There was no error code.

SUnBurst Chart BOWIE STate Enrollment

Source: University System of Maryland, IRIS

Thoughts about SunBurst Chart

I need to type in all my data before I can decide if I like a chart. The Sunburst chart is kind of interesting, it allows for drilling down of data, which I really like. The issue with the particular chart is that there is so much data the chart is kind of difficult to read. In my first version, I did not have the in-State enrollment divided by region and it was even harder to read.

Again I have run into the issue of not being able to get an Amcharts graphic to render on my site. I can make it render in CodePen, but when I export the “pen” it will not render locally. I also can not get it to render using the WordPress plugin for Amcharts. So I am going to park this here until I can figure out why I can not get it to render.

See the Pen amCharts 4: Sunburst Diagram with my data and regions no zero by Caroline (@New-Fish) on CodePen.

Source: University System of Maryland, IRIS

In my previous post, I built my first treemap using Amcharts. Today I built another treemap, this one is 3 levels deep. I am unsure about the presentation of the data in this manner, it may not actually be very useful. However, since I was able to successfully explore building a multi-level treemap using the Amcharts demo I will call it a success.

In this example, I was able to get the level 1 labels to appear, which I was not able to figure out yesterday.

Source: University System of Maryland, IRIS