Kevin Kiley's blog

How to: Learn to develop your own data-based site with no prior knowledge!

How to design a database Web site with limited or no prior knowledge about databases, programming, or Web design.  (Yes, it's possible!)

So, my friend, you found yourself with a massive amount of information that you think your audience would like to see? Want to show the world all those numbers you got through excellent public records requests? Want to get your audience to pick through a pile of facts so you don't have to? Well, then you're going to need a data-driven Web application! And you're going to have to learn how to make one.

Why I hate Ruby on Rails, Mysql, SQLite3 and Windows

 

What has Kevin been doing for the past two months?

I spend most of my time for this class learning how to design a Web application using a programing language called Ruby on Rails. (If you want to see what Ruby is capable of, you can check out these little-known sites.) And as we get down to the last few weeks, I've realized that I still don't have much to show for my efforts. There is no dynamic Web site where we can parse through data. We're still not even sure how to host that information if we build it.

A discouraging meeting and North Carolina's reporting problem

I met with UNC professor Gary Henry in the Department of Public Policy a little more than a week ago. 

He uses a massive database of student statistics (teachers, test scores, race, gender, family, absences) and uses this to measue different policies and how effective they are. He gets a lot of this data through a FERPA exemption, but a lot of it has been built through hard work and the labor of numerous Ph.D. students.

He has worked in multiple states analyzing public education and strategies to lower the dropout rate. And he said North Carolina is a frustrating state to work in.

What we can learn with data (suspensions and the dropout rate)

With all this data that I've been compiling, its hard to find direction. Staring at a spreadsheet of numbers upon numbers, its hard to pick out trends and try and find something meaningful.

And that's why I've been spending so much time trying to build a way to let readers work with the numbers I have in a simple way, so they can look for patterns and trends to help explain this.

Because you have things that you think should be explored. Things that I wouldn't even think of. Take this example:

How the dropout rate for different races has changed over time

This table and graph show the change in dropout rate by race during the past five school years.

The only group to show a dramatic change is American Indians, where the rate has dropped from about 9 percent to about 7 percent. They now show the same rate as Hispanic students, whose rate has not show a discernable change in the past five years. Most other groups seem to follow the same trend as the state, increasing from 2004-2007, and decreasing last year. Each group showed a decrease for the 2007-08 school year, which is something that has been highlighted before.

Finally getting some data

So I finally got some data from the Department of Public Instruction, which will form the backbone to everything else I'm trying to do this semester.

Building a dropout database

My database is growing.

During the past two weeks, I've incorporated data publicly available on the Department of Public Instruction's Web site into a massive collection of facts and figures on each Local Educational Association (LEA). Right now, the database consists of race and gender information, as well as state-mandated testing results.

Ken Gattis with the Department of Public Instruction is working on sending me the dropout information similar to what is available in the annual dropout report, but in a usable form that can be incorporated into the rest of the data.

My pursuit of numbers that mean something

I contacted Ken Gattis at the Department of Public Instruction at the end of last week in hopes of getting some data we could use on this project. For each dropout incident, the Local Educational Association has to list specific characteristics about the dropout, including name, age, gender, race, migrant status, number of grade retentions, number of days missed and the reason for dropping out. To see exactly what each LEA has to fill out, go here and scroll down to page 28 and following.

Some General Assembly bills dealing with education

In the course of my research, I came across a few education bills entering the N.C. Legislature that I thought I should share. Some of them relate directly to the dropout rate, and some of them have a more tenuous relationship with the rate, but could still influence it.

House 65: Students under 16 may attend community college (In committee on Education, if favorable, appropriations).

House 161: Require six-year-olds to attend school. (In committee on education, if favorable, appropriations). This drops the required age from seven to six.

The resources just around the corner

One of the best things about being a reporter at a university is the abundance of experts that we pass every day. Multiple UNC faculty members are researching the issues surrounding the dropout rate, and they could prove an invaluable resource as we move deeper into our exploration.

I've been told that if I want to look at educational policy research, there are a few people I have to go to. 

In the School of Education, I've been pointed to Judith Meece and Lynne Vernon-Feagans.

Syndicate content