We have all heard of IQ—but what about the importance of SQ and EQ? Join SmarterServices and Nuro Retention to learn more about how your students’ social and emotional non-cognitive data directly impacts student success and educational outcomes. Nuro Retention will share how to make BIG data actionable by combining the power of SmarterMeasure Learning Readiness Indicator's non-cognitive data along with its retention software platform and predictive analytics models.
In addition, Dr. Mac Adkins, CEO and founder of SmarterServices, will share a case study on how Ashford University has been able to improve retention rates using the power of non-cognitive data. Nuro Chief Data Scientist Natalie Young will also share some key findings from a recent predictive analytics model that dramatically improved retention efforts for one of Nuro’s clients.
Don’t miss out on your chance to learn the latest strategies on the power of predictive, proactive, and prescriptive data!
Measures of Dispersion and Variability: Range, QD, AD and SD
The Role of Non-Cognitive Indicators in Predictive and Proactive Analytics: Turning Data into Doing
1. Nuro + Smarter Measure’s Combined Power provides
the most comprehensive predictive and proactive
student success solution available on the market today.
SQ/EQ: NON-COGNITIVE
INDICATORS IN STUDENT SUCCESS
WEBINAR AUGUST 7TH @ 2PM CST
2. Founder and CEO of SmarterServices.
Since 2002 he has lead the company as
it has grown to serve over five million
students and twelve thousand
faculty from over five hundred
educational institutions. He served as a
higher education leader for over
twenty-five years and served as a
Director/Dean of Distance Education
for ten of those years. During his career
in higher education, he has also served
as a Director of Enrollment
Management, Director of Student
Services, Director of Instructional
Design, and Data Analyst in the
Department of Institutional Research.
Dr. Mac Adkins Ms. Amy Sorter
Chief Data Scientist for Nuro.
Responsible for building a sophisticated
analytics platform to guide the
development of applications to
empower frontline stakeholders, engage
students, and improve retention.
Catalyst for change, breaking through
conventional wisdoms to champion
transformational programs – ability to
gain executive sponsorship, readily
engages others and cultivates
collaborative relationships, achieves buy-
in and support among key stakeholders,
and advocates for results. Specialties:
CRM, Data Mining and Analytics,
Customer Acquisition and Retention,
Market Research, Brand Positioning,
Marketing Measurement, Consultative
Selling, Employee Development
Ms. Natalie Young
National Sales Director for Nuro.
Ed Tech Software Sales Specialist who
consistently exceeds quota. Marketing
and Sales Manager/ Business
Development with over 20 years of
Marketing, Sales and Relationship
Management experience working to
successfully promote profitability of
each organization. Over 18 years of
experience working in Higher Ed, with
a concentration in Financial Aid,
Enrollment, Retention and Data
Analytics. Passionate about solving
business problems on campuses and
streamlining processes to make the
student lifecycle successful.
Our Presenters
3. Student Success Management
OUR AGENDA TODAY:
The Higher Education Landscape
and Shift Towards Student Success Outcomes
The Need for Comprehensive Solutions
Impact of Non-Cognitive Factors
like SQ/EQ & Predictive Analytics
Combined Power of Nuro + Smarter Measure Solution
Questions and Discussion
1
2
3
4
5
4. • Low Graduation Rates:
The 6 year overall graduation
rate in the US is only 54%.
• Declining Enrollment:
Drop of 2.5M since 2010.
• Regulatory Environment:
Increasing pressure on traditional
colleges to improve performance
• Financial Pressure:
Declining revenues combined
with increasing costs are driving
many colleges to the brink
State Funding
Dependent on
Outcomes
Tuition
Discounting
Better
informed
Students
Adverse
Demographic
Trends
Gainful
Employment
Expansion &
Other Regulatory
Changes
Declining
Enrollment &
Graduation
4
Today’s colleges are facing a challenging environment that
requires them to improve performance to survive.
As retention increases, completion, and graduation rates also increase
ensuring college survival by driving better outcomes and revenue growth.
Colleges are battling for survival
5. v
2018 Bill & Melinda Gates Foundation
45% OF TODAY’S STUDENTS ARE NON-TRADITIONAL
Today’s student is more complex
6. It’s expensive to lose a student!
$9,740
$15,420
$3,530
$0
$2,000
$4,000
$6,000
$8,000
$10,000
$12,000
$14,000
$16,000
$18,000
PUBLIC 4 YR PRIVATE 4 YR PUBLIC 2 YR
Net Tuition Revenue per Full-Time Equivalent Student (FTE)
7. 7That equates to $140MM in annual endowment!
Student population: 10,000
Tuition: $10,000
1st year Retention Rate: 60%
Graduation Rate 50%
Annual Retention Improvement 1%
1ST YR RETENTION RATE 61% 62% 63% 64% 65% 66% 67%
RESCUED STUDENTS 40 117 228 339 450 561 673
ADDITIONAL REV ($M) $0.4 $1.2 $2.3 $3.4 $4.5 $5.6 $6.7
Additional Tuition Revenue (Cumulative)
Just a 1% improvement in retention rates each year results in significant financial
benefit through additional tuition revenue and an increasing number of graduates.
XYZ UNIVERSITY
The Impact of Improved
Student Retention is Massive
8. 8
Highly predictive persistence models
combine pre and post enrollment data
with Smarter Measure non-cognitive
factors to identify
at-risk students starting on day one.
Our software monitors student
performance and engagement throughout
the entire student journey and alerts
frontline stakeholders if students deviate
from their expected paths.
Experienced education professionals
provide high-touch client support to
interpret analytical results and create
action plans to improve performance.
Nuro merges a monitoring software platform, predictive analytics, and experienced
education professionals to deliver an unrivaled student success solution.
CLIENT SUCCESS
& ENGAGE
INSIGHTS &
DISCOVERY
CASE
MANAGEMENT
INSIGHTS & DISCOVERY
CLIENT SUCCESS & ENGAGE
CASE MANAGEMENT
The Solution: Nuro + SmarterMeasure
9. Smarter Services Overview
Founded 2002
600 client institutions
SMARTERMEASURE
Over 4.5 million students have
taken the SMLRI
SMARTERASSESSMENTS
Recently published assessments:
CBERI, TWTRI, Test Anxiety
SMARTERPROCTORING
Only multi-modal proctoring
management system
Partner with ProctorU & Examity
SMARTERID
Only platform for learner
authentication and attendance
during instruction
10. Predicting Student Retention
• Research conducted by Stephen
Nettles, Loraine Devos-Comby, &
Chris Wang at Ashford University
• Research results reported at
the WSCUC-ARC Conference
• Client since 2011
Predicting Student Retention
Utilizing Demographic and
Non-Cognitive Data
11. Predictive Analytics
• Increasing use in higher education
• Classify students / predict student behavior-PREDICTIVE
• Can be used to guide student support - PROACTIVE
12. How?
• Uses existing institutional data to better understand
patterns of student performance & inform decisions
in response to risks
• Pre-matriculation data for early targeted support
• Typically: student demographics and prior achievements
13. Study: Two Sources of Data
• Student Demographics available at pre-matriculation
• SmarterMeasure, an assessment of online learning
readiness (http://www.smartermeasure.com/)
14. Pre-matriculation Student Characteristics
• Number of Transfer Credits
• Employment Status
• Military Status (Active Duty, Veteran, Other)
• Marital Status
• Having Dependents
• Race
• Gender
• Father’s Education
• Age
15. SmarterMeasure Scales
SCALE MEASURES
Life Factors
• Availability of time to study
• Availability of a dedicated place to study
• Reason for continuing one's education
• Support resources from family, friends, and employers
• Perception of academic skills
Personal Attributes
• Procrastination
• Time management
• Persistence
• Willingness to ask for help
• Academic attributes
• Locus of control
Learning Styles • Visual; Verbal; Social; Solitary; Physical; Aural; Logical
16. SmarterMeasure Scales
SCALE MEASURES
Reading Skills
(2 scales)
• Reading rate – Word Per Minute (WPM)
• On-screen reading recall
Technical Knowledge
• Technology usage
• Technology in your life
• Technology vocabulary
• Personal computer/Internet specifications
Technical Competency
• Computer competency
• Internet competency
Typing Skills
(2 scales)
• Typing rate - Word Per Minute (WPM)
• Typing accuracy
17. Population & Methods
• Historical data: students with 1 year of data
• 2,875 Bachelor’s students
• Students tracked weekly (course attendance)
• One-year retention = active at week 52
18. POLL
Using demographic & SM data, what % of
students could you correctly classify
(retained/not) at the end of a student's first yr?
19. Analyses
• First analysis: Based exclusively on demographics
• Second analysis:
Based on demographics AND SmarterMeasure
• Compared prediction accuracy of each analysis
to examine improvement in accuracy from adding
SmarterMeasure in the model.
20. Modeling
• Data divided equally into training and test datasets
• Eight classification algorithms were used:
o Logistic Regression
o Neural Network
o Bayesian Network
o Decision List
o C5.0 Tree
o CHAID Tree
o C&R Tree
o Quest Tree
21. Model Selection
• Forecasting risk to drop out
• For each analysis, the three most accurate algorithms
selected for an optimal ensemble model
• Combined models: Better sensitivity & specificity
than any of the individual models
• Each risk score = average of three algorithms
22. Risk Scores
• Each student was assigned:
o A risk score based on the analysis
with the demographics only
o A risk score based on the analysis combining
demographics and SmarterMeasure
23. Risk Groups
• Students were ranked based on their risk to drop out
and were divided into 10 equal risk groups
(percentiles; N = 287 in each group).
• Model accuracy was computed for the 20% most at-risk students.
24. Model Accuracy
50%
60%
70%
80%
90%
Demographics Only Demographics +
SmarterMeasure
85.50%
SmarterMeasure
Added value: 3.8%
Classification accuracy for the 20%
students with the highest scores:
Dropout Rate for 20% Students With Highest Scores
25. Comparison to average Retention by Risk group
-20
-15
-10
-5
0
5
10
15
20
25
30
35
1 2 3 4 5 6 7 8 9 10
• The retention rate of Group 1 exceeded the average
retention rate by more than 30 percentage points.
• The rates for groups 1-5 were all above the average, while
the rates for groups 6-10 were all below the average.
27. Demographic Results
10% Least At Risk 10% Most At Risk
Older (~40) Younger (~29)
51.2% Married 77.1% Single
Most White Most Non-White
85.4% Female 57% Female
10.8% Veteran
1.7% Active Duty
0.7% Veteran
14.3% Active Duty
50.7% Employed Full Time 65.9% Most Not Employed
67.9% 0-23 Transfer Credits 100% 0-23 Transfer Credits
28. Age
0%
2%
4%
6%
8%
10%
12%
14%
22 24 26 28 30 32 34 36 38 40 42 44 46 48 50 52 54 56 58 60 62 65 78
10% Least At Risk 10% Most At Risk
Age:
Dotted lines indicate the average for each group
10% Least At Risk 10% Most At Risk
Mean = 40.55
SD = 10.33
Mean = 29.34
SD = 6.62
F (1, 578) = 243.37, p < .001; η2 = .30 (large effect)
35. SmarterMeasure Results
10% Least At Risk
10% Most At Risk Scored LOWER than
10% Least at Risk on:
Personal Attributes • Procrastination
• Time management
• Persistence
• Willingness to ask for help
• Academic attributes
• Locus of control
Technical Knowledge • Technology usage
• Technology in your life
• Technology vocabulary
• Personal computer/Internet specifications
Reading Recall • On-screen reading recall
Technical Competency • Computer competency
• Internet competency
Typing Skills • Typing rate - Word Per Minute (WPM)
• Typing accuracy
Learning Styles • Visual; Verbal; Social; Solitary; Phys.; Aural; Logic.
36. PERSONAL ATTRIBUTES
10% Least At Risk 10% Most At Risk
Mean = 83.61
SD = 6.03
Mean = 74.08
SD = 6.58
F (1, 578) = 329.86, p < .001; η2 = .36 (large effect)
0%
2%
4%
6%
8%
10%
12%
10% Least At Risk 10% Most At Risk
38. READING RECALL
10% Least At Risk 10% Most At Risk
Mean = 77.04
SD = 14.07
Mean = 53.41
SD = 20.44
F (1, 578) = 261.94, p < .001; η2 = .31 (large effect)
0%
5%
10%
15%
20%
25%
30%
35%
0 10 20 30 40 50 60 70 80 90 100
10% Least At Risk 10% Most At Risk
39.
40. TYPING SPEED
10% Least At Risk 10% Most At Risk
Mean = 23.06
SD = 11.91
Mean = 14.93
SD = 9.86
F (1, 577) = 80.27, p < .001; η2 = .12 (medium effect)
0%
1%
2%
3%
4%
5%
6%
7%
8%
9%
10%
0 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 42 44 48 51 53 56 59 82
10% Least At Risk 10% Most At Risk
Words/min.
41.
42.
43. Interpretation
• 7 out of 9 SM scales associated with risk groups
• SM improved prediction accuracy by 3.8%
• Substantial overlap between student characteristics and SM;
SM most likely mediating impact of student characteristics on retention
Characteristics
Smarter
Measure
Retention
44. Interventions
• Student characteristics = not malleable factors
• SmarterMeasure as locus of intervention
Remedial strategies could help students overcome
the socio-economic, cultural, and psychological
barriers that put specific categories of students at
risk for poor retention and limited success.
45. Examples of Interventions
SmarterMeasure Scale Interventions
Personal Attributes • Time Management
• Resilience Training
• Learning strategies
• Locus of control
Technical Knowledge • Computer Technology Training
Reading Recall • Speed reading training
• Memory training
Technical Competency • Computer technical skills training
Typing Skills • Typing progrma
Learning Styles • Development of multiple learning styles
46. Retention Management
An educationally sound approach to student success
Early Identification:
• Non-cognitive assessment
- Identifies strengths
- Identifies challenges
- Identifies receptivity to assistance
Early interventions:
• Predictive analytics
• Proactively reaching out to students
according to areas of self-reported
need and risk analysis
• Prescriptive deployment
of intervention strategies
Early
Identification
Early
Intervention
Student &
Institutional
Success
47. 47
Highly predictive persistence models
combine pre and post enrollment data
with Smarter Measure non-cognitive
factors to identify
at-risk students starting on day one.
Our software monitors student
performance and engagement throughout
the entire student journey and alerts
frontline stakeholders if students deviate
from their expected paths.
Experienced education professionals
provide high-touch client support to
interpret analytical results and create
action plans to improve performance.
Nuro merges a monitoring software platform, predictive analytics, and experienced
education professionals to deliver an unrivaled student success solution.
CLIENT SUCCESS
& ENGAGE
INSIGHTS &
DISCOVERY
CASE
MANAGEMENT
INSIGHTS & DISCOVERY
CLIENT SUCCESS & ENGAGE
CASE MANAGEMENT
The Solution: Nuro + SmarterMeasure
48. There is extensive data throughout a student’s journey to predict those at-risk.
In one case, 40% of first term student drops were derived from just 10% of the population!
Predictive models are most impactful when you apply them at the beginning of the student
experience—improving recruitment, enrollment, retention, and ultimately completion.
48
The Power of Predictive Data
49. • Integrated solution to make
predictive indicators actionable
• Middleware technology consumes
multi-source data to provide
stakeholders with analytics to
identify risk, inform intervention
efforts, and provide leadership with
insight into functional areas
• Surveys in conjunction with student
success platform better supports
positive student outcomes leading
to degree completion
•
Consumes and combines cognitive & non-cognitive data
points to make it actionable with Smarter Services Surveys
Nuro Discovery –
Smarter Services Non-Cognitive Risk Score
51. Questions
and Next Steps
Thank you for joining us today!
+
Amy Sorter
National Sales Director, Nuro
asorter@nurolearning.com
813-404-0315
Dr. Mac Adkins
Smarter Services, CEO
mac@smarterservices.com
334.491.0416