Transcript Teaching Stat I - University of Idaho

Teaching Stat I: Putting it all
together
Kirk Steinhorst and Carolyn Keeler
University of Idaho
1
Kirk Steinhorst and Carolyn Keeler
2
The Statistics Teaching
Reform movement
• The problem: Cobb JSE 1993
• Progress: Keeler and Steinhorst JSE
1995
• Current status: Garfield, Hogg, Schau,
and Whittinghill JSE 2002
3
The Essential Ingredients
• Statistical thinking vs. data analysis vs
mechanics
• Technology…statistical computing, webbased learning materials, WebCT
• Activity-based learning and authentic
assessment
4
Course goals-Statistics and
Probability
• Statistical literacy
• Probability as needed in other courses
• Statistical numeracy (ability to calculate
things and draw graphs
• Ability to apply statistics in professional
life and in public life
Kirk Steinhorst and Carolyn Keeler
5
6
Format
• Large lecture, stadium seating, computer
projection and overheads
• 30 minute mini-lecture, 20 minute in-class
group activity
• Instructor can call up web text and applets in
class
• WebCT used for out of class activities (2)
7
Assessment: Activities,
homework and tests
8
Assessment: Activities,
homework and tests
Kirk Steinhorst and Carolyn Keeler
9
The one BIG idea
10
Outcomes
• Grade distribution
– A 29
– B 31
– C 24
– D 10
– F
4 + 8*
Overall average 2.5
– Early dropouts 20 (16% of 126)
11
Midcourse Evaluation +
12
Midcourse Evaluation -
13
Diagnostic quiz-weeks 1 & 14
14
End of Course Evaluation +
15
End of Course Evaluation -
16
Survey of Attitudes Towards
Statistics (Candace Schau
Univ New Mexico)
• 28 Likert items reduce to 4 scales –
– affect (+ - feelings) 6 items
– cognitive competence (intellectual skills) 6 items
– value of statistics 9 items
– difficulty of statistics 7 items
• 115 pre, 72 post, 64 complete change scores –
a decrease of about 5 points in the first 3 scales
and about 3 points on the difficulty scale
17
Remember our plan…
Use active learning in cooperative groups
with web-based instructional materials
and authentic assessment
Kirk Steinhorst and Carolyn Keeler
18
Findings
• The system works for the top half of the
class
– Those students who use the technology
like it.
– In-class active learning works.
– In-class tests based on takehome tests
leads to effective authentic assessment.
19
Findings
• The poor or below average student
has great difficulty with “statistical
thinking” – technology doesn’t help.
– WebCT based discussions were not
successful.
– The poor student could not use the webbased text effectively.
– Poorer students need regular homework
in addition to activities.
20
Recommendations
• Use a conventional textbook with on-line
applets
• Assign and collect homework regularly
• Use Minitab or SYSTAT
• Keep in-class activities and takehome/inclass testing
21
Conclusions
1. Active learning in cooperative groups
can work in large classes
2. Use web-based instructional materials
carefully –too much technology
overwhelms the poorer student
3. Activities and tests as described are
good models for authentic assessment
Kirk Steinhorst and Carolyn Keeler
22
Final comment on the plan:
Use active learning in cooperative groups
with web-based instructional materials
and authentic assessment…
We went a step or two too far. Back off on
web text and web discussion. The rest is
good.
23