Skip Navigation

A Conversation About High School Dropouts and Completers.

Chat Host Hi, as you may have learned from reading the report, the analysis of high school dropouts and high school completers is a rather complex undertaking. That is why I am hosting this chat to field and answer some of your questions. My hope is that by answering one person's question and then widely sharing the answer, I can clarify how the study was conducted for a large group of people. That being said, let's get started.

Paul Gammill from Crofton Md asked:
Chris, The definition of dropout and completer is more complex than a lay person would expect. Can you point me to an NCES definition of these two terms? Do all SEA?s and LEA?s use this same definition? Because the term dropout has political implications, how confidant are you that this data is being reported accurately?
Chris Chapman: This is a good question with which to start. Definitions of dropouts and completers can be found in appendix C of the "Dropout Rates in the United States: 1999" report. All SEA's (state education agencies) and LEA's (local education agencies) in states for which we have data in the state-by-state dropout rate table (table 2) use the Common Core of Data (CCD) definition of who is a dropout. You can also find a detailed definition of CCD's reporting instructions to state CCD Coordinators at:, a pdf of the exact instructions is at the bottom of the page.

The state-level dropout data that we do have appear to be sound. If you roughly average out the reported state rates and compare that against the national dropout rate from the Current Population Survey (CPS), the estimates are roughly in line with one another at around 5%. The CPS is a good bench mark against which to compare because it is based on responses from individuals and not various education agencies. We will be able to do a more detailed subnational comparison once the decennial Census data are released. Also, based on response patterns from the SEAs and LEAs, we do have confidence that their reports are accurate. Unfortunately, because this is a short answer forum, I can't go into more detail here on this issue.

Lida from Flagler Beach, FL asked:
How do we motivate underachievers to want to stay in school? What is being taught (or the way curriculum is presented) really doesn't seem relevant to many students. Is there any major change in the works as far as secondary education in the state of Florida is concerned?
Chris Chapman: The annual dropout rate report that we put out is really designed to provide an annual snapshot showing how well we are doing in terms in decreasing the number of dropouts and increasing the number of high school completers. However, there are some good sites that have information on dropout prevention programs.

1. National Dropout Prevention Center

2. National Institute on the Education of At-Risk Students

3. Summaries of research from the U.S. Department of Education

Probably the best place to start in terms of seeing what changes might be in the works for Florida would be your state education agency.

Lucy Hood from San Antonio, Texas asked:
If I understand the numbers correctly, the completion rate has been steady in recent years, but there has been an increase in the number of those receiving a GED. Does that mean there is a decrease in the number of students receiving a high school diploma, and if so, do you know why? Are high school graduation requirements more stringent, and if so, is that a factor? Thank you, Lucy Hood, reporter with the San Antonio Express-News.
Chris Chapman: You did read the report correctly. This is what we did report out. There is indeed a decrease rate of those receiving a regular high school diploma. We have not done research looking at why this might be occurring, however.

Robert from Little Rock Arkansas asked:
What does your study say about the influence of # of hours worked on drop-out and academic performance? [ controlling for SES?]
Chris Chapman: At the moment, we do not study the influence of hours worked on school performance in the dropout report although this is an idea that might be worth pursuing further. However, findings from other reports do suggest a significant correlation between hours worked and dropping out of school. For a good example of such a study using Center data, I would recommend, "A Comparison of High School Dropout Rates in 1982 and 1992" at:

Patrick from Augusta, Maine asked:
Why aren't GED completers included in the completion rate calculation?
Chris Chapman: This relates back to an earlier question about the complexity of the rates that we do report. In fact, in our national level dropout and completion estimates based on the Current Population Survey (CPS) data, GED recipients are counted as completers. In "Dropout Rates in the United States: 1999" all of the completion rates use the CPS data and therefor include GED recipients.

Unfortunately, on the Common Core of Data (CCD) some state and local education agencies have difficulty tracking who earned GEDs in their jurisdictions meaning that databased on agency reports lack GED counts. The completion rate is calculated using district-level dropouts and completers. There are two issues with GEDs: one, GEDs are only reported to the CCD at the state level and therefore are not in included in the district-level calculation and two, not all states report GED's on the CCD so the across state comparison would be questionable if they were included. A technical report was released in February of this year detailing the methodological work that went into creating this rate and NCES's recommendations for the completion rate, it can be found on the web at:

Jonni from Portland, Oregon asked:
Is "English as a Second Language" one of the categories of the NCES dropout study? If so, how does the dropout rate among ESL students compare with that of English native language students - regardless of race or ethnicity? Also, if you did categorize by native language, how many years back are there statistics for the dropout rate among ESL/LEP students? How has this changed over the years you have categorized for native language? Thank you!
Chris Chapman: We do not have questions about ESL enrollment in the data collections used to study dropouts. However, we do ask questions about English language proficiency on an intermittent basis. The last time we reported these data out was in the 1995 dropout report which you can find at :

We collected these data again last year and hope to be able to report out the English language proficiency tests sometime next year. Until we have had a chance to look at the 1999 data, we won't be able to say much about trends.

Michael from Alexandria, VA asked:
Is there any statistically significant trend in either the dropout rate(s) or the completion rate? The conclusion of the report says both have remained "stable" in the 1990s, but also says there has been a "gradual pattern of decline" in the status dropout rate over the past 28 years. From a statistical standpoint, can there be both an overall decline and a periodic stability? Thanks...
Chris Chapman: The overall trend from 1972 to the present has shown a decrease in the dropout rate and an increase in the completion rate. However, most of this change occurred before the 1990s. In other words, yes, there can be an overall decline and periodic stability.

Sandra Svoboda from Toledo, OH asked:
By not using the cohort rates, does the extent of dropouts get under-reported? Is cohort rate not used in this national look because of the complexity of collection? Because states use different rates?
Chris Chapman: This is a good question. We don't believe that our dropout rates are too small based on cross sectional data collections. The cohort rate is not used more regularly because it is based on longitudinal data which are indeed more complex and difficult to collect. When we start our next longitudinal study of youth, we will report out new cohort rates.

Chris Ottenbreit from Cheney Washington asked:
What is the drop out rate of middle school students nationwide and in the state of washington?
Chris Chapman: The CCD collects dropout information for students in grades 7-12 so you could use the CCD to calculate a dropout rate of any combination of these grades for any one state. NCES does not have a data collection that collects dropout data at the school level, the CCD collection is at the district level and can be aggregated to the state-level. Also, since not all states report their dropout data to the CCD, using the NCES definition, we can not get national totals using the CCD.

Richard from Portland, OR (Northwest Regional Educational Lab) asked:
The Digest of Education Statistics includes a table that shows high school graduates as a percent of the 17-year-old population (Table 104 in the 1999 Digest). Why aren't these data used in the Dropout Rates report and would you warn against using this measure when calculated at the state level?
Chris Chapman: We do not include this table in the dropout report ,at the moment, because we already include a wide range of indicators that use many different calculations. The Digest table uses a method of calculation different than any of the other tables in the report. To cut down on possible confusion, we do not include the Digest table.

Bob from Salem, Oregon asked:
How do we make comparisons between the event dropout rates NCES calculates and the status dropouts rates based on the current population survey? Where do the two statistics overlap? Where do they pat company?
Chris Chapman: The event rate shows how many youth dropout between the beginning of the school year of the year before and the beginning of the current school year. The status rate tells us how many youth have dropped out irrespective of when they dropped out.

If your question is about differences between Common Core of Data and Current Population Survey estimates, please see appendix C of "Dropout Rates in the United States: 1999" at:

Have you [or anyone] done a similar study on COLLEGE dropouts and completers
Chris Chapman: We do not have similar estimates of status or event dropout estimates for postsecondary students. However, you can calculate cohort dropout rates by using some of our longitudinal data. In particular, Beginning Postsecondary Students Longitudinal Study (BPS) can be used to develop such an indicator. We have something along these lines in the Digest of Education Statistics table 315 at:

Patrick from Augusta, Maine asked:
Is it appropriate to take 100% minus the completion rate and call it the cohort dropout rate?
Chris Chapman: Not really. The cohort rate that we discuss is designed to measure the dropout rate of a particular sample of individuals at multiple points in time or longitudinally. For instance, we started tracking a group of 8th graders in 1988 and just finished another follow-up with those individuals this year. Their dropout rate would be considered a cohort dropout rate.

100% minus the completion rate also does not provide an comparable estimate to the status rate because the completion rate looks at 18-24 year-olds and the status rate is based on 16-24 year-olds. We start at 18 for the completion rate because we know that most people graduate high school by age 18. We use 16 for the status rate because we want to be able to detect if people are leaving high school before they would likely graduate (by age 18).

Lucy Hood from San Antonio, Texas asked:
Once again another question. The report says the completion rate in 1972 was 83 to 84 percent. Now it is 85.5 percent. That means, I assume, it has increased only slightly in 28 years????
Chris Chapman: You interpretation is correct. There has only been a slight increase in the rate.

A. Sultan from St. Louis, Mo asked:
How has the black-white gap narrowed in terms of drop-outs, if the dropout and completion rates have remained constant over the past several years?
Chris Chapman: The gap has actually narrowed since the early 1970s, but has remained relatively constant since the late 1980s. In other words, the narrowing of the gap has paralleled the trends in the overall completion rate patterns. For more detail on the issue, take a look at the overview essay of the 1999 Condition of Education at:

vanessa from new york, NY asked:
Has the NCES done a regression analysis of factors that affect the high school dropout rate?
Chris Chapman: We do not do multivariate analysis in the dropout report because it is designed to provide basic indicators of dropouts. However, regression analysis could be conducted. It is a good question.

Ralph Smith from Austin, TX asked:
I have been asked to construct a crosswalk between NCES/CCD leaver codes and those used by my state education agency. By any chance, has NCES already done this work on a state-by-state basis?
Chris Chapman: Very good question. We have not done this work yet. However, if you are interested in talking with us about how best to go about doing the crosswalk, please contact Beth Young at:

Michael from Alexandria, VA asked:
The conclusion of the report suggests the stable dropout rate may be due to an "increased use of alternative methods of high school completion". Any data to back that up, either in the report or elsewhere? Thanks.
Chris Chapman: The most conclusive data that we present supporting this statement can be found in table 6 of the "Dropout Rates in the United States: 1999" report. More detailed data can be found in appendix C of the report if you are interested.

Alia from Tempe, Arizona asked:
I noticed your state-by-state table of dropout rates went only to 1997-98. Is this what you based your results on, or are there available numbers for 1998-99. I am particularly interested in where Arizona ranks.
Chris Chapman: Data from the Common Core of Data (CCD), the data used to calculate the state level dropout rates, are 1 year behind the Current Population Survey (CPS) data because of the reporting time frames. Next year's dropout report will release the 1998-99 state level dropout data from CCD.

James from Starkville asked:
How long have you been reporting dropouts? Do you also provide any projections of what you think the dropout rates will be in the future? for individual states?
Chris Chapman: We have been analyzing dropout data from as long ago as 1972. Before point in time, we lack information needed to provide the kind of detailed distributions shown in our annual reports.

Lucy Hood from San Antonio, Texas asked:
Me again. One more question. Do you know how many students are represented by the 79.2 percent who completed high school, how many of those received a diploma and how many received a GED? Thanks again.
Chris Chapman: I believe your question refers to the most recent three year Texas completion rate average in table 5. If this is correct, we do not have large enough sample sizes to allow us to calculate stable estimates of the number of people in any given state earning a GED.

Susie from NY, NY asked:
The report concludes that increasing use of alternative methods to graduation like GED pursuit might be having a stabilizing effect on the dropout rate. What were the various alternative routes looked at in the report?
Chris Chapman: Technically, the question we ask has a response category that states "high school graduate-high school diploma or the equivalent (for example GED)". It is possible that we are capturing some other alternative diplomas and equivalencies using this response category.

Kyle from Denver, Colorado asked:
What are the factors that contribute to making a student "at-risk" of dropping out? Where would you suggest I look for more detailed information on those factors?
Chris Chapman: Take a look at the response to Lida's question for some useful sources of information on this topic. OK, time for one more.

Doug from San Jose, California asked:
I really like the report that you just put out. I especially like having an html version in addition to a pdf file. Should I expect that you publish a similar report this time every year?
Chris Chapman: This is a good way to end the chat talking about next year's report. This is great idea, and I hope that we can continue to provide the reports in multiple formats. We do plan on publishing this report every November at least for the near future.

Thanks for all of the good questions, some of them will definitely have an impact on what goes into next years report. Unfortunately, I could not get to all of the questions, but you know how to reach me in case I missed yours. I hope that you found this session to be helpful and the report to be interesting.