Teaching Statistics in the Online Environment

1
Teaching Statistics in the Online Environment

• Michelle Everson, Ph.D.
• Department of Educational Psychology
• University of Minnesota
• gaddy001_at_umn.edu

2
Overview

• Course development and structure
• Focus on active learning/conceptual understanding
  of course material
• Description of student assignments and
  assessments
• What do students think about the course?
• Some lessons learned
• Future directions

3

• I really like the online format and I appreciate
  the individual attention that I receive from the
  TA and instructor. It's a little funny that an
  online course feels more personal than an actual
  in-person lecture.
• The instructor for this course has made me feel
  comfortable if/when I have any questions/concerns,
  and has fostered a true learning environment in
  which the ultimate goal is understanding the
  content.
• (Comments from EPSY 5261 students)

4
The GAISE Recommendations

• According to the Guidelines for Assessment and
  Instruction in Statistics Education (GAISE,
  2005), introductory statistics courses at the
  college level should
• Emphasize statistical literacy and develop
  statistical thinking
• Use real data
• Stress conceptual understanding
• Foster active learning
• Use technology
• Integrate assessments that are aligned with
  course goals
• Source http//www.amstat.org/education/gaise/

5
Research on Teaching Statistics Online

• How can collaborative activities and technology
  can be integrated into an online statistics
  course?
• Student-to-student interaction and collaboration
• Weekly chats (e.g., Dereshiwsky, 1998)
• Project work (e.g., Davis Chao, 2004 Prater
  MacNeil, 2002 Suanpang, Petocz, Kalceff, 2004)
• Regular group discussions (e.g., Grandzol, 2004
  Jones, 2003 Tudor, 2006)
• Technology
• Courses have used Excel, SPSS, Minitab,
  Cyberstats, and ActivStats (e.g., Davis Chao,
  2004 Dutton Dutton, 2005 Grandzol, 2004
  Harrington, 1999 Lawrence Singhania, 2004
  Mills Xu, 2005 Prater MacNeil, 2002 Tudor,
  2006 Utts et al., 2003 Zhang, 2004)

6
Introductory Statistical Methods (EPSY 5261)

• This is a 3-credit, semester-long, graduate-level
  introductory course taught through WebCT
• Prerequisite is high-school algebra
• Students in the course come from a variety of
  different areas
• College of Education
• School of Nursing
• Other programs Journalism, Speech
  Communication, Business, etc.
• Students take the course for a variety of reasons
• Some plan to go on to other courses
• Some will not need to take any other statistics
  courses
• Many select the online course because of
  convenience

7
(No Transcript)
8
(No Transcript)
9
Assignments and Assessments

• Grades are based on
• Small-group Discussion Assignments (8)
• Homework Assignments (4)
• Critique of an Academic Journal Article (1)
• Quizzes (3)
• Project (1)
• Final Exam (1)
• Students also have the opportunity to complete
  non-graded practice activities and extra credit
  assignments

10
Collaborative Discussion Assignments

• Each student is assigned to a discussion group at
  the beginning of the semester
• Eight small-group discussion assignments are
  completed
• Assignments involve discussing concepts and
  answering questions as a group
• Students must post their own thoughts AND respond
  in a meaningful way to what at least one group
  member has posted
• One student volunteers to lead each discussion
  and submit a summary to the instructor by
  midnight on Monday

11
Assignment 1 Designing the Perfect Experiment

• Students are told the story of Scott Ginsberg
  (see the website www.hellomynameisscott.com)
• Students discuss how they might design an
  experiment to determine if wearing a name tag
  causes others to be more friendly (or causes a
  person to become more approachable)
• Through discussion, students talk about issues
  such as sampling, bias, control, and other
  experimental design issues

12
Assignment 4 Sampling Distributions

• Students work independently through a lab in
  which they use the Sampling SIM program (delMas,
  2001)
• Students then attempt to answer a series of
  questions about the lab as a group
• They talk about what they feel a sampling
  distribution is and why it is important.
• They talk about why they think the sampling
  distribution has the characteristics that it
  does.
• They work as a group to answer a question that
  involves applying knowledge of sampling
  distributions.

13
Assignment 7 Correlation and Regression

• Students talk about what variables might predict
  success in college, and how we might measure
  success
• After initial discussion, students are asked to
  examine a real data set that includes test
  scores, GPAs (high-school and college), and other
  data from several college students
• Students try to determine, based on the data set,
  how they might construct a regression equation to
  predict college GPA
• Through discussion, students talk about their
  understanding of both correlation and regression

14
(No Transcript)
15
(No Transcript)
16
(No Transcript)
17
Homework

• Students complete 4 homework assignments, each
  worth 10 points
• Assignments involve answering questions from the
  course textbook
• Most assignments involve using SPSS or other
  technology (e.g., Java applets)
• Assignments are submitted as Word attachments
  through WebCT e-mail
• The TA grades each assignment and sends
  individual feedback to the student within one
  week

18
Project

• Each student completes a project that involves
  gathering data from two groups and
  describing/analyzing the data using SPSS
• The project is submitted in parts
• Part 1 Project idea
• Part 2 Project data
• Part 3 Introduction and description of data
• Part 4 Inference (confidence intervals and
  hypothesis testing) and summary/conclusion

19
Exams

• Quizzes are administered (roughly every 4 weeks)
  online through the WebCT quiz tool
• Available from noon on Fridays until noon on
  Mondays
• Students have up to three hours to complete the
  quiz (in one sitting)
• Quizzes consist almost entirely of short-answer
  questions
• Students complete a take-home final exam
• The final is a project that involves answering
  questions about a data set using SPSS
• The BIG question How do you administer quizzes
  online and ensure that students are not cheating?
  Can you do this????

20
Preventing Academic Dishonesty in the Online
Environment

• If all (or most) of your quizzes will be online,
  you can do different things to prevent academic
  dishonesty
• Change assessments from semester to semester
• Insist that students complete the quiz in one
  sitting, within a certain period of time
• Use open-ended questions where students must
  explain their answers
• You can also randomize the order in which
  questions are presented to each student
• Ask that students adhere to an honor code of
  some kind

21
Student Feedback

• In midterm feedback surveys, students were asked
  to indicate how strongly they agreed that certain
  components of the course contributed to their
  understanding of statistics
• Data from the last three semesters of EPSY 5261
  were pooled
• Overall, student feedback (both at midterm and at
  the end of the semester) has been relatively
  positive

22
More Student Feedback
23
Student Performance

• Students who take this course tend to be highly
  motivated, and most do well (the typical course
  grade is an A)
• Although we have not systematically compared
  performance in this online course to a
  traditional face-to-face course, we have noticed
  that students tend to do about the same
  regardless of the format of the course
• Certain online students, however, might benefit
  more from a traditional setting

24
Lessons Learned

• Teaching online can be a big time commitment
• Online courses are NOT for everyone!
• You get to know your students in a much different
  way when teaching online
• Students appreciate timely communication with the
  instructor/TA, organization, and consistent
  deadlines
• Online discussion assignments can be a great way
  to learn more about student difficulties/misconcep
  tions
• They motivate ALL students to participate
• Discussions can be monitored for ALL GROUPS from
  start to finish
• Set deadlines discourage students from waiting
  until the last minute to participate

25
Future Directions

• In the future, we hope to spend time
• Developing software tutorials to help students
  learn to use SPSS
• Comparing the online course to a face-to-face
  counterpart
• Learning more about how students think and reason
  about statistics by looking carefully at the
  quality of discussions
• Determining how to better motivate students
  (especially undergraduates) to participate more
  actively in the online course

26

• Thank you!!!