Predict Factor Descriptions

Factor: Age

Models: Persistence, Course Completion

Feeds: Person, AcademicTerm

We calculate the age of the student from the person feed birthdate field at the census date of the term (12% of the way through the term as calculated by the academic calendar feed). We then split each of the students into three even bins based on the age of everyone enrolled in the term.

 

Factor: Attempted Credit 

 

Models: Persistence, Course Completion 

 

Feeds: Course, AcademicTerm

We take the sum of all course credits for all non-remedial courses within the current term, and then group students into three even groups based how many credits they are attempting in that term.

 

Factor: Course Level


Models:Course Completion


Feeds: Course, AcademicTerm

We use the first digit in the three digit course code to identify the course level. Levels are assigned as 0; 1; 2; 3 and 4; or 5 and higher

 

Factor: Course Load


Models: Persistence,

Course Completion


Feeds: Course, AcademicTerm

We take sum of all non-remedial courses attempted within the current term, and then group students into three even groups based on the number of courses they are attempting in that term regardless of the credits involved.

 

Factor: Course Modality


Models:

Course Completion


Feeds: CourseSection

Flagged based on the course_online field within the CourseSection data feed

 

Factor: Cumulative Credit Completion Ratio


Models: Persistence, 

Course Completion


Feeds: Course, AcademicTerm

We take the count of all courses taken for credit and the the number of courses where the quality points earned was greater than zero. This gives us the ability to take the ratio of the count of all courses where quality points were earned versus all attempted courses. Remedial and transfer courses are not included.

 

Factor: Cumulative English GPA


Models: Persistence, 

Course Completion


Feeds: Course, AcademicTerm

We calculate GPA using the sum of quality points earned against the sum of total attempted credit for english courses. English courses are found using a lookup to the department name and acronyms. This only includes terms that ended before the start of the current term, and does not include any remedial or transfer courses

 

Factor: Cumulative GPA


Models: Persistence, 

Course Completion


Feeds: Course, AcademicTerm

We calculate GPA using the sum of quality points earned against the sum of total attempted credit. This only includes terms that ended before the start of the current term, and does not include any remedial or transfer courses

 

Factor: Cumulative Math GPA


Models: Persistence, 

Course Completion


Feeds: Course, AcademicTerm

We calculate GPA using the sum of quality points earned against the sum of total attempted credit for math courses. Math courses are found using a lookup to the department name and acronyms. This only includes terms that ended before the start of the current term, and does not include any remedial or transfer courses

 

Factor: Cumulative SAP Flag


Models: Persistence, 

Course Completion


Feeds: TranscriptCourse

We take count of all courses taken for credit and the number of courses where the quality points earned was greater than zero for every term in their past. This gives us the ability to take the ratio of the count of all courses where quality points were earned versus all attempted courses. If it below a 67% credit completion ratio, then they get an Active Cumulative SAP flag. If it isn't true now, but was true when looking at prior terms, then they get a flag for the Prior Cumulative SAP flag. Remedial and transfer courses are not included.

 

Factor: Dependents at Home


Models: Persistence, 

Course Completion


Feeds: PersonFafsa

Using the person_fafsa table, we use the boolean field of dependents_other_than_spouse. We use only the most recent fafsa submitted

 

Factor: Effective Family Contribution


Models: Persistence, 

Course Completion


Feeds: PersonFafsa

Using the person_fafsa table, we use the numeric field of expected_family_contribution. We then bin students with this data available into the top 50% and bottom 50% of students to determine high and low efc. We use only the most recent fafsa submitted

 

Factor: Enrolled in Spring


Models: Persistence, 

Course Completion


Feeds: AcademicTerm, CourseSection

If a student is in their first enrolled term at the institution, and it is a spring term, then they are marked as a New Spring Student. This is only active during spring terms. Transfer credits do not count as previous enrollments here

 

Factor: Enrollment Group


Models: Persistence, 

Course Completion


Feeds: PersonDegreeProgram

We set a threshold of number of attempted credits in a term (6 for two year institutions, 12 for four year institutions). If a student attempts fewer credits than that threshold, then they are counted as part time. The degree seeking status is taken from whether they have a listed degree actively in the degree_name field of the person_degree_program feed

 

Factor: FAFSA Submission


Models: Persistence, 

Course Completion


Feeds: PersonFafsa, AcademicTerm

If a student has a person_fafsa record associated with the current academic calendar year, then they are given the fafsasubmit_current tag. If they have a prior person_fafsa record, but none associated with the current year, then they are given a not_current flag. If they have no person_fafsa records, then they are given a never submitted flag. This grouping is only available to institutions with active fafsa feeds enabled

 

Factor: Gender/Minority


Models: Persistence, 

Course Completion


Feeds: Person, Race, Ethnicity

We use the person feeds coming into SSE, and group students by M and F genders along with the minority field coming from the person feed. This gives us four categories, one for each combination of gender and minority values. If a person does not identify in a minority race or ethnicity group, they are grouped into the non-minority group. Students without data are also grouped into the non-minority group due to them not explicitly identifying in a minority group

 

Factor: Has Withdraw


Models: Persistence, 

Course Completion


Feeds: TranscriptCourseStatus

We use the transcript_course data feed to find any previous terms where the status is "Withdrawn", "web_withdrawal", "-", "#", or "*"

 

Factor: Has Math Withdraw


Models: Persistence, 

Course Completion


Feeds: TranscriptCourseStatus, Course

We use the transcript_course data feed to find any previous terms where the status is "Withdrawn", "web_withdrawal", "-", "#", or "*". This only includes math courses, as identified by their department name and acronym

 

Factor: Has Remedial Withdraw


Models: Persistence, 

Course Completion


Feeds: TranscriptCourseStatus, Course

We use the transcript_course data feed to find any previous terms where the status is "Withdrawn", "web_withdrawal", "-", "#", or "*". This only includes developmental/remedial courses, as identified by the is_developmental tag on the course feed

 

Factor: Math Course


Models: 

Course Completion


Feeds: Department, CourseSection

Course is identified as a math course through its department name or acronym. Only designated math departments are included.

 

Factor: Transfer Student


Models: Persistence, 

Course Completion


Feeds: Student, AcademicTerm, CourseSection

We identify transfer credits three ways.
Using any prior enrollments with Transcript Course data associated with a Transfer Credit AcademicTerm.
Using the CourseSection src_system_id to find courses tagged as 'tran'
Using the CourseSection transfer field as a transfer flag.
If a student has any enrollments identified in any of these ways, they are marked as a transfer student

 

Factor: Marital Status


Models: Persistence, 

Course Completion


Feeds: PersonFafsa

Identified through the dependent status feed using Dependent/Rejected-Dependent or the Independent/Rejected-Independent values.

Independent students then we use the student's marital status from the person fafsa feed
Dependent students we use the parent's marital status from the person fafsa feed

 

Factor: Previous A Grade


Models: Persistence, 

Course Completion


Feeds: TranscriptCourse

Looks at all non-transfer, non-remedial courses where transcript_course.final grade = "A"

 

Factor: Previous F Grade


Models: Persistence, 

Course Completion


Feeds: TranscriptCourse

Looks at all non-transfer, non-remedial courses where transcript_course.final grade = "D" or "F"

 

Factor: Previously Taken


Models: 

Course Completion


Feeds: Course, TranscriptCourse, AcademicTerm

Repeat courses are found using the Course feed and the TranscriptCourse feed. A course is considered repeated if a student has the same course with the same course Id in their transcript history. That course counts as failed if it is a non-transfer, non-remedial course where transcript_course.final grade = "D" or "F". That course counts as withdrawn if that transcript_course_status is "Withdrawn", "web_withdrawal", "-", "#", or "*"

 

Factor: Previous Summer Enrollment


Models: Persistence, 

Course Completion


Feeds: AcademicTerm, TranscriptCourse

Using all registered TranscriptCourse records, a student is identified as having a prior summer enrollment if they were enrolled in one or more courses during a summer term prior to this current term

 

Factor: Prior Enrollments


Models: Persistence, 

Course Completion


Feeds: AcademicTerm, TranscriptCourse

The count of the number of academic terms where a student has a transcript course record is taken, and then students are grouped into 0, 1, 2, and 3+ groups of terms enrolled.

 

Factor: Program Enrollment Count


Models: Persistence, 

Course Completion


Feeds: PersonDegreeProgram

Count of the number of actively enrolled programs based on the PersonDegreeProgram feed

 

Factor: Program Path Risk


Models: Persistence, 

Course Completion


Feeds: PersonDegreeProgram

We evaluate the historical risk of each program at the institution using the percentage of students enrolled in each program who had successful persistence and course completion outcomes. We then rank the programs into low, medium, and high risk paths based on that percentage using the bottom third as high risk, the next third as medium risk, and the top third as low risk. Once we have the relative risk of each program, we then assign students as low, medium, and high risk based on their enrolled program. If a student is enrolled in multiple programs, they are given the risk level of their highest risk program.

 

Factor: Attendance Alert


Models: Persistence, 

Course Completion


Feeds: Attendance, TranscriptCourse

Using the attendance feed, if enabled, students are flagged if they have the first_week_attendance_alert_generated field marked.

 

Factor: Registration Timing


Models: Persistence, 

Course Completion


Feeds: TranscriptCourse, TranscriptCourse-StatusHistory

Using the first course registration as the 'registered date' of each student for each term, we then split the students into the top half and bottom half of registration times, to see who registered for their first course of the term earlier compared to their peers

 

Factor: SAP Flag


Models: Persistence, 

Course Completion


Feeds: TranscriptCourse, AcademicTerm

When the total number of credits earned is fewer than 67% of the credits attempted for any prior enrollments at the institution, then a student is marked with the SAP flag. This only looks at each term individually, not the cumulative credits across terms

 

Factor: Transfer Credit Count


Models: Persistence, 

Course Completion


Feeds: TranscriptCourse, AcademicTerm

Sum of the total Credit Earned associated with Transfer Course Section records

 

Factor: Transfer Year Count


Models: Persistence, 

Course Completion


Feeds: CourseSection, AcademicTerm

Counts the number of academic years a student was enrolled in transfer courses, based on the academic term year field associated with each transfer course section

 

 

 

 

 

Articles in this section

How to Contact Support
There are many ways to reach out! Click here for our support options.
Watermark Academy
Click to access the Watermark Academy for consultation, training, and implementation companion courses.
Watermark Customer Community
Engage and connect with others!