The Transformative Power of Remediation

1
The Transformative Power of Remediation

• Steven Spurling
• City College of San Francisco

2
The Questions

• Are English and Math levels related to success in
  a wide variety of college (IGETC) academic areas?
• If so, does remedial course-taking in English and
  mathematics allow students who start in college
  at lower levels to perform equally to students
  who start at much higher levels?

3
The dataset

• 10 Years of Academic History
• 88,000 Students
• 350,000 IGETC Course Enrollments
• 20 IGETC Areas
• Nearly a Million Class Meetings
• Untold Student Time Investment
• Native-Speaking Students Only NO ESL
• Intersegmental General Education Transfer
  Curriculum (IGETC)

4
Two Hypothetical Patterns

• The Student Learning Outcomes Pattern
• The Academic Talent Pattern
• We will sift through the dataset looking to
  determine which of these two patterns explains
  student success.

5
The Student Learning Outcomes Pattern -
Hypothetical
6
The Academic Talent Pattern
7
The Actual Pattern
8
Actual Number of Students
9
The Data Layout
10
IGETC Areas
11
CCSF Sequences
12
Zero Order Correlations
13
Regression Estimates allowed entry into a
stepwise regression at a .0001 level
14
GPA and N in English Composition
15
Summary Effects Sorted by Impact
16
Sorted By Relative Math/English Importance
17
Two Major Questions

• How do students do once they get to where ever
  they are going educationally?
• How many of them get there?
• The first question needs to be answered before
  the second one has any meaning.

18
Question 2 The Completion of Transfer-Course
Rate of New First Time Students
19
Are the progressively lower completion rates for
lower placing students in the prior slide a
result of lower placing students being filtered
out because of their lack of academic capability
or because they are required to take more class
levels?

• To Complete a transfer level
• Arithmetic students must
• Enroll in arithmetic
• Pass it
• Enroll in elementary algebra
• Pass it
• Enroll in intermediate algebra
• Pass it
• Enroll in a transfer level math class
• Pass it
• An EIGHT step process

20
Whats P?

• How to compare the ability to complete the
  sequence of courses of higher placing and lower
  placing students?
• The march to D.C. analogy.
• A simplifying assumption all of the percentages
  are the same
• p the average enrollment and success rate.
• T transfer level completion
• S steps in the sequence
• S (l 1)2 1
• Where l levels below transfer
• t ps
• p t 1/s

21
Whats P in English?

• P (L) 11th root of .0348 .739
• P (90) 9th root of .153 .812
• P (92) 7th root of .242 .817
• P (93) 5th root of .356 .813
• P (96) 3rd root of .498 .793

22
And 8 Years Out?
23
ITS THE LENGTH OF THE SEQUENCE!!!

• P is very similar across disciplines and from one
  level to the next.
• Students at the lower levels are not less likely
  to succeed and re-enroll at any given point.
• Their lower completion rates are related to the
  number of steps they have to complete! Its the
  length of the sequence!

24
Conclusions

• Academic Talent or Learning Outcomes?
• It would seem to be Learning outcomes however in
  English there is grade slippage that impacts
  performance exterior to English.
• Social promotion or the soft C or both?
• Sequence length or filtering of less capable
  students who enter the sequence at lower levels?
• The exponent is the length of the sequence.
• P represents the equality of capability across
  levels.
• It would seem to be sequence length (the
  exponent) that lowers end-of-sequence completion
  more than differences in P average
  point-in-time success and enrollment rates by
  level.
• P declines markedly at the lowest levels only in
  a four year examination. In an 8 year look, P
  levels out across levels.
• The Bottom-bottom line.
• We can deal with many issues. These include
  everything from learning outcomes to sequence
  length to grade slippage to heterogeneity of
  ability. We cant affect academic talent.
• Given the nature of our populations, low placing
  and part-time, we need to track them over longer
  periods to time. Our model might be continuous
  education over long periods of time.
• The transformative power of remediation or the
  second press? Whats our job?

25
Transfer Completion by Levels below Transfer and
Average Success and Enrollment Rate