1. Pathways to Learning in Undergraduate Statistics
Joe Ferrandino, Ph.D.
Associate Professor of Criminal Justice
School of Public and Environmental Affairs (SPEA)
Indiana University Northwest

2. The Connection between LEAP Goals and Undergraduate Statistics is Very Strong
Knowledge of Human Cultures and the Physical and Natural World
Through study in the sciences and mathematics, social sciences, humanities, histories, languages, and the arts
Intellectual and Practical Skills, Including
Inquiry and analysis
Critical and creative thinking
Written and oral communication
Quantitative literacy
Information literacy
Teamwork and problem solving
Personal and Social Responsibility, Including
Civic knowledge and engagement—local and global
Intercultural knowledge and competence
Ethical reasoning and action
Foundations and skills for lifelong learning

3. Some Important Context
54% of US students are unprepared for college math (ACT, 2012)
Students:
Fear math in general and statistics specifically
Come in unprepared
Are anxious
Often drop the course, withdraw or fail (high DFW)
Put off taking the class until the very end
A total of 151 students enrolled in course over my 4 previous semesters:
17 failed, mainly by not completing the course (11%)
12 withdrew sometime after the first week (7.9%)
13 dropped before add/drop ended
Lost 42 of potential 164 students (25.6%)

4. Philosophical and Pedagogical Change: My Two Cents Worth
Philosophy
Embrace learning statistics through failure
Refocus the orientation from performance to achievement
Unlimited opportunity through student determination that teaches me about students while they learn
Connect students together and make them responsible for each other’s learning and success

5. Pedagogy
Regardless of modality, students have unfettered access to video lectures and assignment tutorials
All quizzes and tests would have unlimited opportunities for students to fail, learn from the failure, and succeed (Learning through failure as a means to an end and no longer and end in itself; testing for learning)
Achievement would replace performance (minimum standards set without limits on reaching 100%; 70% on exams and 80% on quizzes to continue in course)
Determination to learn and succeed would become a key component of the course going forward (eliminate excuses and learn about student achievement)
Peer-based learning through discussion board questions and answers

6. The Structure, Visually, of the Old and the New
Assignment
Tutorial

7. The Initial Results
I present here the initial results from two different courses (both undergraduate 300 level statistics courses)
Spring 2015: 15 week hybrid course with 48 students
Summer 2015: 6 week online course with 100 students
These results include:
student retention, completion,
Learning, and
what I learned about students

8. Retention and Completion
One goal of the change in this course was to increase student retention and completion in addition to improved learning and enhanced overall experience. Recall, the “old course” had an established DFW rate of 25.6% (42 of 164 total enrolled students did not complete the course successfully). In the spring hybrid course, the DFW percentage was reduced to 14.6% (7 of 48 enrolled students) and remained 14% (14 of 100) in the summer 6-week fully online course.

9. Student Learning

10. A Single Quiz: Actual Student Learning Curves

11. Student Persistence, Determination and Achievement
Traditional testing method (where every student takes each assessment once):
this would equate to 2,193 attempts (129 students each taking 17 assessments) if no one missed a single attempt. The previous semester before making this change, students missed 21.2% of all attempts, not including students that withdrew.
In the previous semester before making the change, 178 of the 217 possible attempts students could have taken on the quizzes (82%) were below the newly established 80% threshold.
New testing method (unlimited attempts)
Produced a total of 9,477 attempts, or a minimum 332% increase over the traditional, singular high stakes method.
Even with the low stakes of unlimited opportunity testing, the percentage of quiz attempts below the threshold to continue in the course was 52.7% (3,552 of 6,742 quiz attempts) and eventually reduced to 0 based on the course structure.
Just 246 attempts (3.6%) were students achieving a perfect score on their first attempt, showing the depth of student need for multiple opportunities. In fact, when given the opportunity to continue working and learning, a total of 789 other quiz attempts (11.7%) eventually reached the perfect score level, with students requiring an average of 4.1 attempts to achieve that result (NOTICE NOT EVERYONE STRIVES FOR A 100!)
It also must be noted that initial quiz scores in both revised courses was similar to the initial performance of the class in the previous semester, whose performance averaged between 4.93 and 5.42 on 7 comparable quizzes that could only be taken once.

12. Learning About Students and their Learning
I learned that most students overachieve
if given the chance (and also underperform
if allowed to). Minimum standards are essential.
One test was given without the required 70%
score to continue: 12 of 38 students with more
than one attempt (31.6%) scored below the
minimum 70% threshold and stopped taking the
test. Paradoxically, the test had the most attempts
with the least overall achievement

13. Conclusion
Sometimes, changes cannot be made at the margins of a course
I turned failure into a strategy that eased anxiety, placed the onus on students to be determined and exert more effort, and improved their opportunities to learn
I increased collaboration between the students as the course moved online (or hybrid)
I focused on achievement rather than performance
The course has been a much better experience overall, except for the students that do not want to put in the work or the effort