WEBVTT 00:00:05.046 --> 00:00:06.036 I'm Elaine Allensworth. 00:00:06.346 --> 00:00:10.886 I'm the Co-Director for Statistical Analysis at the Consortium on Chicago School Research 00:00:10.966 --> 00:00:13.276 at the University Of Chicago in Chicago, Illinois. 00:00:13.926 --> 00:00:17.516 A data system that follows students' records as they move 00:00:17.516 --> 00:00:20.866 through high school is essential for measuring dropout rates. 00:00:21.166 --> 00:00:24.716 If you can't follow students from the time they start in high school 00:00:24.746 --> 00:00:26.306 to the time they leave high school, 00:00:26.626 --> 00:00:30.656 you will not accurately estimate the dropout problem in your school. 00:00:31.126 --> 00:00:35.386 Longitudinal data systems that follow students as they move from year to year 00:00:35.426 --> 00:00:39.526 to year is especially critical for understanding the nature of the dropout problem 00:00:39.966 --> 00:00:43.546 because dropout is a process that occurs over a number of years. 00:00:44.106 --> 00:00:47.756 There are many different kinds of data systems that schools and districts 00:00:47.756 --> 00:00:50.226 and even states are using right now. 00:00:50.466 --> 00:00:55.976 Traditionally, if there have been data systems, they've been school based or district based. 00:00:56.216 --> 00:00:59.256 Increasingly, we're seeing more and more state-based systems. 00:00:59.256 --> 00:01:02.586 If we can link records from across schools within a district, 00:01:03.006 --> 00:01:07.126 we can tell if a student has transferred to another school within the district rather 00:01:07.126 --> 00:01:10.206 than just wondering if the student has transferred or if they've dropped out. 00:01:10.636 --> 00:01:14.586 If we can follow students as they move from district to district within the state, 00:01:14.986 --> 00:01:17.236 that takes away more of the question 00:01:17.236 --> 00:01:19.726 of whether students have transferred, whether they've dropped out. 00:01:20.086 --> 00:01:27.166 When you're designing the systems, you always have to balance the complexity of the system 00:01:27.766 --> 00:01:32.496 with the logistics and the actual staffing that you're going to have 00:01:33.076 --> 00:01:35.706 to make sure that that system runs well. 00:01:36.096 --> 00:01:44.126 You're going to have to always make tradeoffs in terms of thinking what will be the easiest 00:01:44.286 --> 00:01:48.956 to implement and think about what are the demands on the clerks, 00:01:49.256 --> 00:01:54.336 what are the demands on the teachers, what are the demands on central office administration 00:01:54.336 --> 00:01:58.446 who needs to process the data, on the demands of the technology folks? 00:01:58.686 --> 00:02:03.736 It's a challenge putting together statistics that makes sense and answer the questions 00:02:03.826 --> 00:02:09.866 that school staff really have about why students are struggling, why they are dropping out. 00:02:09.866 --> 00:02:13.486 A lot of times, you can start with some very basic kinds of questions: 00:02:13.526 --> 00:02:17.136 What are the dropout rates by student's incoming test scores? 00:02:17.196 --> 00:02:22.296 What are the dropout rates by student's age at entry into high school, Special Education status? 00:02:22.446 --> 00:02:23.086 Things like that. 00:02:23.086 --> 00:02:26.656 And that can start giving you some idea of the scope of the problem 00:02:26.976 --> 00:02:29.096 and the structure of the problem in your school. 00:02:29.446 --> 00:02:31.736 But then the question is what more? 00:02:31.936 --> 00:02:35.106 What do we do with the data from there? 00:02:35.106 --> 00:02:39.816 What we find is, generally, the more that you can make comparisons across schools, 00:02:40.086 --> 00:02:43.786 across sub-groups of students and across schools, and even across districts, 00:02:44.236 --> 00:02:49.816 the more you get insight into the nature of the problem in your particular school. 00:02:50.456 --> 00:02:53.666 To develop early warning indicator systems, 00:02:53.836 --> 00:02:59.096 the most crucial information that schools need is information on students' grades 00:02:59.376 --> 00:03:01.016 and their attendance in their classes. 00:03:01.416 --> 00:03:04.326 A lot of times, there's an assumption that students drop 00:03:04.326 --> 00:03:07.356 out because they have weak academic skills. 00:03:07.756 --> 00:03:11.356 But in reality, students drop out because they've stopped coming to class, 00:03:11.666 --> 00:03:15.766 because they're not putting in the effort, because they failing their classes 00:03:15.896 --> 00:03:20.056 and as a result of failing their classes, they're realizing that they're not going to graduate. 00:03:20.776 --> 00:03:27.226 Once the data system is in place, it takes time and it takes effort to develop people's capacity 00:03:27.306 --> 00:03:33.076 to look past the raw numbers to be able to identify patterns in what's happening in the school. 00:03:33.266 --> 00:03:38.656 But that's really critical because once you start looking at which students are failing, 00:03:38.986 --> 00:03:44.016 why they're failing, when they're failing, when absence is worse, when absence is better, 00:03:44.346 --> 00:03:48.706 that's when you can start making strategic decisions about how the policies 00:03:48.706 --> 00:03:51.476 and practices in your school and in your district are working 00:03:51.526 --> 00:03:54.916 and you can start really addressing the problems of dropout 00:03:55.036 --> 00:03:59.976 and making a difference for your kids.