The Most At-Risk First Time in College Students: Using Assessment Techniques to Identify and Assist Them Prepared for the International Assessment and Retention Conference, Assessment Reconsidered: Improving Learning, Student Success, and Transparency, 11–15 June 2008. Scottsdale, Arizona. Bernadette M.E. Jungblut, Ph.D. – Assistant Director Daniel Suleski – Coordinator of Statistical Research Brittany L. Resmann – Coordinator of Assessment and Research Office of Assessment and Planning Student Development and Enrollment Services University of Central Florida Orlando, FL 32816-0160
The Most At-Risk First Time in College Students: Using Assessment Techniques to Identify and Assist Them. Prepared for the International Assessment and Retention Conference, Assessment Reconsidered: Improving Learning, Student Success, and Transparency , 11–15 June 2008. Scottsdale, Arizona. - PowerPoint PPT Presentation
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
The Most At-Risk First Time in College Students: Using Assessment Techniques to Identify and Assist Them
Prepared for the International Assessment and Retention Conference, Assessment Reconsidered: Improving Learning, Student Success, and Transparency, 11–15 June 2008. Scottsdale, Arizona.
Bernadette M.E. Jungblut, Ph.D. – Assistant DirectorDaniel Suleski – Coordinator of Statistical ResearchBrittany L. Resmann – Coordinator of Assessment and Research
Office of Assessment and PlanningStudent Development and Enrollment ServicesUniversity of Central FloridaOrlando, FL 32816-0160
Workshop Goals
Discuss the methods used to predict which First Time in College (FTIC) students are the most likely to be at risk for retention failure in their first year at UCF
Describe a targeted intervention, the Knight Success Program, designed to assist these students and increase the likelihood of their engagement and persistence
Provide session participants with several opportunities to: Discuss the applicability of these methods at your institutions Discuss the generalizability of this type of intervention to your
institutions Examine the efficacy of this approach in your specific institutional
contexts
Workshop Agenda
Introduction
Short activity to assess participants’ needs and primary interests
UCF Project Overview
Discussion of applicability to participants’ institutions
UCF Project Findings and Next Steps
Discussion of participants’ expectations about at-risk students
UCF Program Components
Discussion of program’s generalizability to other institutions
Short Activity: Questions to Consider Are you currently identifying at-risk incoming students?
If so, what techniques are you using to identify those students? Who are your at-risk incoming students?
If not, do you plan to do so in the future – and what factors do you expect to be “likely suspects” to help you identify these at-risk students?
Do you currently have targeted intervention programs in place for at-risk incoming students? If so, choose one (1) to focus on for this activity. How is this program structured? What are the program’s primary components? How is this program staffed? How are you recruiting students for this program? How are you connecting and communicating with these students?
If you do not have targeted intervention programs in place for at-risk incoming students, are you planning to design and implement such programs? What components would you include in such a program?
Facts about UCF
Largest undergraduate enrollment and second largest total enrollmentin Florida
Sixth largest university in the country by total enrollment; second largest in the country in terms of undergraduate enrollment
Knight Success Program – DevelopmentHypothesis: The university can enhance its first-year retention rate by engaging
students who are most at-risk for dropping out after their first year.
Action Plan: Develop a program to engage our most at-risk students.
Problem: How do we know which students are the most at-risk before they begin
their college career?
Knight Success Program – DevelopmentSolution:
Use what we already know about our incoming students to formulate a predictor of first-year retention.
Resources available:
Information new students provide to admissions
Information current and former students have provided to admissions, along with their retention outcomes
Knight Success Program – Data
What does the university know about students before they arrive on campus?
Standardized test scores High school grades and coursework attempted Class rank and percentile Financial aid On-campus housing requests Personal demographics (age, gender, ethnicity, etc.) Residency (in-state, out-of-state, international) Name and type of high school (public/private/home school/etc.) College credits earned (Advanced Placement, Dual Enrollment,
International Baccalaureate, etc.) Major student is pursuing (or undeclared status)
Factors to Examine at Your Institution Which of these factors are relevant to your
institution?
Which are not relevant or applicable?
What other factors would you need or want to include in your models?
Knight Success Program – Data Analysis Model building: Use information from past admissions and outcomes to
construct, test, and validate predictive models.
Model selection: Choose the model with the highest hit rate.
Hit rate: Total correctly predicted non-retained students divided by the total number of students not retained
Knight Success Program – Modeling
Entropy split criterion found to be the best model
Knight Success Program – Decision Rules Decision Rule 1:
- High school GPA less than 3.25- About 30% of 1,000 most at-risk students
Decision Rule 2:- Total SAT score is greater than or equal to 1240- High school GPA is between 3.25 and 4.15- Student is a National Merit Scholar Finalist
- About 10% of 1,000 most at-risk students
Decision Rule 3:- Total SAT score is greater than or equal to 1000- SAT math score is greater than or equal to 460- High school social sciences unit GPA is greater than 4.05- High school English unit GPA is greater than 4.75
- About 60% of 1,000 most at-risk students
Knight Success Program – Results Retention Outcomes
KSP-selected students have a lower retention rate than non-selected students.
The KSP model is an effective predictor of at-risk students.
KSP participants have a higher retention rate than non-participants.
KSP participation is effective in improving first-year retention.
KSP Modeling Efficacy
75.9%
81.8%82.7%
84.1%
70.0%
73.0%
76.0%
79.0%
82.0%
85.0%
2005/2006 2006/2007
Year Enrolled/Retained
Re
ten
tio
n
KSP Selected KSP Not Selected
KSP Program Efficacy
81.1%82.6%
74.7%
79.3%
70.0%
73.0%
76.0%
79.0%
82.0%
85.0%
2005/2006 2006/2007
Year Enrolled/Retained
Re
ten
tio
n
KSP Participant KSP Non-Participant
Other Student Outcomes of KSP Academic Probation
Students selected for KSP are typically more likely to be placed on academic probation than non-selected students.
Participants in the KSP program are less likely to be placed on academic probation than non-participants.
Probation
0.0%
4.0%
8.0%
12.0%
16.0%
20.0%
Summ
er 2
005
Fall 2
005
Sprin
g 200
6
Summ
er 2
006
Fall 2
006
Sprin
g 200
7
Summ
er 2
007
Fall 2
007
Term
Pe
rce
nt
Pla
ce
d o
n P
rob
ati
on
Not KSP 2005
KSP Select 2005
Not KSP 2006
KSP Select 2006
Not KSP 2007
KSP Select 2007
Performance Gap of KSP Selected vs. Not Selected Percent Placed on Probation by Term
0.0%
2.0%
4.0%
6.0%
8.0%
Summ
er 2
005
Fall 2
005
Sprin
g 200
6
Summ
er 2
006
Fall 2
006
Sprin
g 200
7
Summ
er 2
007
Fall 2
007
Term
Pe
rce
nta
ge
Dif
fere
nc
e
Performance Gap2005
Performance Gap2006
Performance Gap2007
Other Student Outcomes of KSP Term GPA
Performance Gap of KSP Selected vs. Not Selected Term GPA
0
0.1
0.2
0.3
0.4
Summ
er 2
005
Fall 2
005
Sprin
g 200
6
Summ
er 2
006
Fall 2
006
Sprin
g 200
7
Summ
er 2
007
Fall 2
007
Term
Dif
fere
nc
e in
Te
rm G
PA
Performance Gap2005
Performance Gap2006
Performance Gap2007
Students selected for KSP typically have lower term GPAs than non-selected students.
Participants in the KSP program typically have higher term GPAs than non- participants.
KSP Selected vs. Not Selected Term GPA
2.6
2.8
3
3.2
3.4
Summ
er 2
005
Fall 2
005
Sprin
g 200
6
Summ
er 2
006
Fall 2
006
Sprin
g 200
7
Summ
er 2
007
Fall 2
007
Term
Pe
rce
nt
Pla
ce
d o
n P
rob
ati
on
Not KSP 2005
KSP Select 2005
Not KSP 2006
KSP Select 2006
Not KSP 2007
KSP Select 2007
Next Steps
Comparison of data mining models with deductively derived logistic regression models
More in-depth examination of additional factors FCAT scores (Florida’s state-wide assessment of high school
students) First generation in college status Socioeconomic factors including financial aid status AP, IB, and Dual Enrollment credits earned in high school High school curricula and credits earned by subject area
CSRDE NSSR Follow-up Presentation: 29 September–1 October
KSP Components Years 1 and 2 Year 1 (AY 2005-2006)
August 2005 start (for September 2005 program launch) Identified 1000 students; recruited 200 active participants Incentives provided (bookstore coupons) Distributed among all 12 first-year advisors Skills assessments for 200 active participants Engagement events for all 1,000 students invited to participate
Year 2 (AY 2006-2007) May 2006 start (for summer and fall program infusion) Identified 1000 students; recruited 200 active participants No incentives provided except opportunity for skills assessment Assigned to one of two dedicated KSP Advisors Infused into approximately 16 Orientations (May through August) Skills assessment = Strength Quest (for 200 students) Engagement events E-mail contact SLS 1501 – Strategies for College Success course
KSP Components Years 3 and 4 Year 3 (AY 2007-2008)
Very similar to Year 2 Same two KSP Advisors Added First Year Advising and Exploration (FYAE) Peer Advisors Special Orientation sessions for KSP selected students (logistical
problems)
Year 4 (AY 2008-2009) Somewhat similar to Year 3 Same two KSP Advisors Focus on engagement and academic success strategies Enhanced communication and contact (e-mail plus more face-to-face) Orientation sessions revised
Day 1 = KSP Cohort special sessions Day 2 = College schedule planning (to address Year 3 logistical problems)
KSP Advisors design 20- to 30-student intensive pilot program
Program Components
Which of these KSP components are relevant to your institution?
Which are not relevant or applicable?
What other components would you need or want to include?