**About the Data**

College and Career Readiness Early Warning System (CCREWS) predictions are available for most students in grades 6 through 9 beginning in 2017-18. CCREWS provides predictions for four college- and career-related student outcomes:

- Whether or not a student will
**enroll in a four-year college**degree program in the first fall after high school graduation - A student’s predicted overall
**ACT® composite score** - Whether or not a student will meet the college readiness benchmark on
**three of the four**ACT® subjects (English, mathematics, reading, science) - Whether or not a student will meet the college readiness benchmark on
**four of the four**ACT® subject

CCREWS predictions are calculated using student-specific data reported by Wisconsin school districts to the Department of Public Instruction. See the CCREWS webpage for more information.

CCREWS is intended to be used as a starting point for inquiry. Schools are encouraged to combine the CCREWS predictions with more current local information and contextual information to consider the case of each student as well as other data available in WISEdash.

In addition to the CCREWS predictions, each student is assigned a level of risk for each of four areas where educators could have an impact: attendance, behavior, mobility, and assessment.

**Definitions**

**CCREWS College Enrollment Prediction:** CCREWS provides a prediction of whether or not a student will **enroll in a four-year college** degree program in the first fall after high school graduation. This prediction is either “Ready” or “Not Ready”.

**CCREWS College Enrollment Probability:** CCREWS provides the probability that a student will **enroll in a four-year college** degree program in the first fall after high school graduation. The probability ranges from 0-100%. This probability is calculated individually for each student and represents how often similar students enrolled in college. Thus, a student with a probability of 60% has reported data that looks very similar to students in previous cohorts who enrolled in college 60% of the time. A probability of 50% or more results in the prediction “Ready” while a probability of less than 50% results in the prediction “Not Ready”.

**CCREWS College Enrollment Lower Bound**: CCREWS provides a lower margin of error that represents the error or uncertainty surrounding the college enrollment probability for an individual student.

**CCREWS College Enrollment Upper Bound**: CCREWS provides an upper margin of error that represents the error or uncertainty surrounding the college enrollment probability for an individual student.

**CCREWS ACT Composite Score**: CCREWS provides a predicted ACT® composite score. The score ranges from 1-36. No predicted outcome of “Ready” or “Not Ready” is provided for the ACT® composite score because ACT® has not established a college readiness benchmark for the composite.

**CCREWS ACT Composite Lower Bound**: CCREWS provides a lower margin of error that represents the error or uncertainty surrounding the composite score for an individual student.

**CCREWS ACT Composite Upper Bound**: CCREWS provides an upper margin of error that represents the error or uncertainty surrounding the composite score for an individual student.

**CCREWS ACT 3/4 Subjects College Ready Prediction:** CCREWS provides a prediction of whether or not a student will meet the college readiness benchmark on **three of the four** ACT® subjects (English, mathematics, reading, science). The prediction is either “Ready” or “Not Ready”.

**CCREWS ACT 3/4 Subjects College Ready Probability**: CCREWS provides the probability that a student will meet the college readiness benchmark on **three of the four** ACT® subjects (English, mathematics, reading, science). The probability ranges from 0-100%. This probability is calculated individually for each student and represents how often similar students met three of the four benchmarks. Thus, a student with a probability of 60% has reported data that looks very similar to students in previous cohorts who met three of the four benchmarks 60% of the time. A probability of 50% or more results in the prediction “Ready” while a probability of less than 50% results in the prediction “Not Ready”.

**CCREWS ACT 3/4 College Ready Lower Bound**: CCREWS provides a lower margin of error that represents the error or uncertainty surrounding the ACT® 3/4 subjects college ready probability for an individual student.

**CCREWS ACT 3/4 College Ready Upper Bound**: CCREWS provides an upper margin of error that represents the error or uncertainty surrounding the ACT® 3/4 subjects college ready probability for an individual student.

**CCREWS ACT 4/4 Subjects College Ready Prediction**: CCREWS provides a prediction of whether or not a student will meet the college readiness benchmark on **four of the four** ACT® subjects (English, mathematics, reading, science). The prediction is either “Ready” or “Not Ready”.

**CCREWS ACT 4/4 College Ready Probability**: CCREWS provides the probability that a student will meet the college readiness benchmark on **four of the four** ACT® subjects (English, mathematics, reading, science). The probability ranges from 0-100%. This probability is calculated individually for each student and represents how often similar students met three of the four benchmarks. Thus, a student with a probability of 60% has reported data that looks very similar to students in previous cohorts who met four of the four benchmarks 60% of the time. A probability of 50% or more results in the prediction “Ready” while a probability of less than 50% results in the prediction “Not Ready”.

**CCREWS ACT 4/4 College Ready Lower Bound**: CCREWS provides a lower margin of error that represents the error or uncertainty surrounding the ACT® 4/4 subjects college ready probability for an individual student.

**CCREWS ACT 4/4 College Ready Upper Bound**: CCREWS provides an upper margin of error that represents the error or uncertainty surrounding the ACT® 4/4 subjects college ready probability for an individual student.

**CCREWS Mobility Risk**: Information is provided about four risk factors that may impact college and career readiness. Mobility is one such factor. Students who did not change schools in the prior year are assigned low risk. Students who changed schools once in the prior school year are assigned moderate risk. Students who changed schools more than once in the prior school year are assigned high risk. Note that the risk factors are not predictions but instead are based on actual student behavior.

**CCREWS Discipline Risk**: Students who were not suspended or expelled from school or suspended or expelled for less than one day are assigned low risk. Students suspended or expelled 1 or more days and fewer than 2.5 days in the prior school year are assigned moderate risk. Students suspended or expelled for 2.5 or more days are assigned high risk.

**CCREWS Attendance Risk**: Students absent for fewer than 12.5 days in the prior school year are assigned low risk. Students absent 12.5 to 20 days in the prior school year are assigned moderate risk. Students missing greater than 20 days of school in the prior school year are assigned high risk.

**CCREWS Assessment Risk**: Students who score lower than 0.75 standard deviation below the median combined English language arts and mathematics score are assigned high risk. Students who score higher than 0.75 standard deviation above the median combined English language arts and mathematics score are assigned low risk. Students that score between are assigned moderate risk. These scale score ranges vary by grade and by year. They are different than proficiency cut points and based on equating the current student’s performance to the performance of students classified in each of the three risk categories.

**CCREWS Production Date:** This is the date the CCREWS predictions were created.

Students who do not receive CCREWS predictions do not have enough data to make the required calculations. This does not mean these students are not college and career ready; rather that a school should investigate why no score is available and determine college and career readiness based on the most current local data.

**How CCREWS is Reported in WISEdash**

The CCREWS predictions are accessible through the Early Warning dashboard in the Advanced Analysis tab of WISEdash for Districts. By navigating to this dashboard and selecting CCREWS in the Early Warning System drop-down menu, school staff can view a student list including CCREWS predictions. More detailed CCREWS information for any individual student can be viewed by accessing the Student Detail view through the CCREWS student list and scrolling down to the Early Warning (DEWS and CCREWS) section.

CCREWS predictions are updated twice a year. At the beginning of each school year, DPI releases preliminary CCREWS predictions for students in grades 6 through 9. When the required data become available final CCREWS predictions replace the preliminary CCREWS predictions. The most recent CCREWS predictions for a student are always displayed.

**Agreement Across CCREWS Predictions**

CCREWS provides predictions for four outcomes for each student. For a given student the relationship between these predictions will often agree with what a educator might expect to see based on their knowledge of ACT® subject benchmarks, ACT® composite, and college enrollment. Despite this, educators should be aware that CCREWS predictions may not always agree with their expectations as the prediction for each outcome is based on a unique model that does not refer to any of the other CCREWS predictions.

For a practical example of the predictions behaving as educators might expect, consider the relationship between the ACT® three out of four subjects outcome and the ACT® four out of four subjects outcome. When we look at the CCREWS predictions we see three groups of students: 1) students who are predicted to be “ready” on both of these outcomes, 2) students who are predicted to be “not ready” on both of these outcomes, and 3) students who are predicted meet the benchmark on **three of the four** ACT® subjects but are predicted to not meet the benchmark on **four of the four** ACT® subjects. These three groups make intuitive sense because it is more difficult for a student to meet all four of the college readiness benchmarks.

Sometimes the relationship between the predictions is not as clear-cut. Educators might intuitively expect that a student who is predicted to meet three or four of the ACT® benchmarks should also be predicted to achieve a high ACT® composite score. Often this will be true but remember that CCREWS predictions are based on how past students performed. There are students in the past who met the benchmark on three subjects and performed very poorly on a fourth, resulting in a relatively low ACT® composite score. In addition, our opinion of whether these predictions agree also depends on what we define as a high score. A student who just meets the benchmark on all four subjects will have a composite score of 21.25 (18 + 22 + 22 + 23 / 4 = 21.25).

College enrollment is probably the least intuitive of the predictions. Educators may expect that students who are predicted to do well on ACT will also be predicted to enroll in college in their first year following high school graduation. Of course, many students who could attend a four-year college instead choose to enter the workforce, join the military, or enroll in a two-year college. Since past students made these choices, CCREWS predicts that some current students will also not enroll in college even though they might otherwise appear to be good candidates for a four-year college degree.

**Agreement Between DEWS and CCREWS Predictions**

The Dropout Early Warning System (DEWS) provides a prediction of whether a student is at risk to drop out before graduation from high school. Specifically, DEWS predicts a student to be at low, moderate, or high risk for dropout. Typically the DEWS prediction and the CCREWS predictions will agree in a way that makes intuitive sense. That said, educators should not expect that a particular DEWS prediction will be associated with any given pattern of CCREWS predictions as the two systems do not refer to each other when making predictions.

Note that the DEWS dashboard student list will likely include many more records than the CCREWS dashboard student list. This is because DEWS has been in operation for longer and shows results for years prior to 2017-18. Similarly, because DEWS and CCREWS are separate systems that are, in part, based on different data and are run at different times, not all students who receive a DEWS prediction will receive CCREWS predictions and vice-versa.

**View more information about CCREWS at the CCREWS Home page.**