SBI listserv participants,
We are happy to announce three new posts for you to enjoy:
- "Teaching computation as an argument for simulation-based inference
<https://www.causeweb.org/sbi/?p=971>" by Mine Cetinkaya-Rundel in our
series on 'Why use simulated based methods?"
- "Reflections after two years of simulation-based inference in AP
statistics <https://www.causeweb.org/sbi/?p=954>" by Andrew Walter in our
series on 'Incorporating simulation-based methods in the HS classroom"
- "Using simulation-based methods in New Zealand
<https://www.causeweb.org/sbi/?p=976>" by Stephanie Budgett in our new
series on "Simulation-based methods around the world"
We hope that these posts will challenge and encourage you. These posts and
more are available at http://www.causeweb.org/sbi
Feel free to continue to suggest other potential topics for the blog, post
your questions/comments directly to the blog and/or send your thoughts
using this listserv.
If you aren't aware of it already, encourage your students with excellent
class projects to consider submitting them to the USPROC competition which
has prizes for all levels of student projects (intro stat and beyond; see
www.causeweb.org/usproc for details)
Nathan
--
Nathan Tintle, Ph.D.
Associate Professor of Statistics and Dept. Chair
Director for Research and Scholarship
Dordt College
Sioux Center, IA 51250
nathan.tintle(a)dordt.edu
Phone: (712) 722-6264
Office: SB1612
Statistics education colleagues,
We currently have two introductory stats courses here at Central College.
* Applied Statistics (MATH 215) is intended for more mathematically/technically mature/advanced students. We use Rossman and Chance's ISCAM book, and students learn how to use Minitab to do statistical tests and intervals.
* Intro to Statistics (MATH 105) is intended for students who are not as sophisticated mathematically/technically. We use Tintle's Introduction to Statistical Investiations book, and students use the accompanying applets to do much of the work for them.
We are struggling with finding an effective prerequisite to use to put students in these two courses. Does anyone have ideas?
The biggest difference between the two courses is how fast and how deep we can go with the material. In Applied Stats, we go faster and make some deeper connections with the material. We also expect that students can fill in some of the gaps as we go along. But in Intro to Stats, we go slower and spend more time filling in gaps and making connections for the students.
We are struggling with finding an effective prerequisite to use to put students in these two courses. In particular, we'd like to have an enforceable prerequisite for each course that would keep good students from just taking the easy road with Intro to Stats. Does anyone have ideas?
Currently, we simply use math placement results to decide this. Students placing at or above Calc I or having completed (at least) Precalculus are not eligible to enroll in MATH 105. However, as strange as this seems, our current registration system cannot check/enforce this prereq, so any student can actually enroll in either course. So we end up having to police these criteria ourselves--usually removing students from Intro to Stats and encouraging them to enroll in Applied Stats. (Not a happy job...)
Lately, we have begun to wonder if using a math placement result that is based on a scale from College Algebra to Multivariable Calculus is really the best way to measure what will make a student successful in a particular stats class.
We have done some analysis of a number of different possible predictors of student success, and the one that seemed to be the strongest was cumulative GPA. A GPA of 2.7 seemed to be the low end for students who successfully completed Applied Stats. So we proposed a pre-req of GPA <= 2.7 for Intro to Stats and GPA >= 2.7 for Applied Stats, but our Registrar doesn't like it and has asked us to consider a different pre-req.
We've talked a lot about this as a department, and we really don't know where to go.
If anyone out there with a similar situation of having two intro courses has an easy, effective, and enforceable way to determine student placement, then I would enjoy hearing from you. Please just reply directly to me and not to the SBI mailing list.
Thanks!
Mark
Dr. MARK A. MILLS
Professor of Mathematics | Central College
812 University Street | Campus Box 06 | Pella, Iowa 50219
Mark-
Thanks for your email. Here's what we're doing here at Dordt---but I, too,
would love to hear more.
1. We have two intro courses. The AP stat equivalent which uses the ISI
book you are using and another 'accelerated' intro stats which does the ISI
book in half a semester and then does a half semester of multivariable
applied statistics.
2. Our placement criteria for the second course is (a) taken calculus (even
though we don't use any), or (b) prior experience with statistics (e.g., AP
statistics or some other college or HS course)
3. We do have some students take the 'easier' intro course instead of the
longer one, but not that many so it's not an issue. When we get overbooked
in the regular intro course these better students are the first to go... :)
Again, curious what others are doing as well.
Nathan
On Fri, Dec 4, 2015 at 2:55 PM, Mark Mills <MillsM(a)central.edu> wrote:
> Statistics education colleagues,
>
>
>
> We currently have two introductory stats courses here at Central College.
>
> - Applied Statistics (MATH 215) is intended for more mathematically/
> technically mature/advanced students. We use Rossman and Chance's
> ISCAM book, and students learn how to use Minitab to do statistical tests
> and intervals.
> - Intro to Statistics (MATH 105) is intended for students who are not
> as sophisticated mathematically/technically. We use Tintle's *Introduction
> to Statistical Investiations* book, and students use the accompanying
> applets to do much of the work for them.
>
> We are struggling with finding an effective prerequisite to use to put
> students in these two courses. Does anyone have ideas?
>
>
>
> The biggest difference between the two courses is how fast and how deep we
> can go with the material. In Applied Stats, we go faster and make some
> deeper connections with the material. We also expect that students can
> fill in some of the gaps as we go along. But in Intro to Stats, we go
> slower and spend more time filling in gaps and making connections for the
> students.
>
>
>
> We are struggling with finding an effective prerequisite to use to put
> students in these two courses. In particular, we'd like to have an
> enforceable prerequisite for each course that would keep good students from
> just taking the easy road with Intro to Stats. Does anyone have ideas?
>
>
>
> Currently, we simply use math placement results to decide this. Students
> placing at or above Calc I or having completed (at least) Precalculus are
> not eligible to enroll in MATH 105. However, as strange as this seems, our
> current registration system cannot check/enforce this prereq, so any
> student can actually enroll in either course. So we end up having to
> police these criteria ourselves--usually removing students from Intro to
> Stats and encouraging them to enroll in Applied Stats. (Not a happy job...)
>
>
>
> Lately, we have begun to wonder if using a math placement result that is
> based on a scale from College Algebra to Multivariable Calculus is really
> the best way to measure what will make a student successful in a particular
> stats class.
>
>
>
> We have done some analysis of a number of different possible predictors of
> student success, and the one that seemed to be the strongest was cumulative
> GPA. A GPA of 2.7 seemed to be the low end for students who successfully
> completed Applied Stats. So we proposed a pre-req of GPA <= 2.7 for Intro
> to Stats and GPA >= 2.7 for Applied Stats, but our Registrar doesn't like
> it and has asked us to consider a different pre-req.
>
>
>
> We've talked a lot about this as a department, and we really don't know
> where to go.
>
>
>
> If anyone out there with a similar situation of having two intro courses
> has an easy, effective, and enforceable way to determine student placement,
> then I would enjoy hearing from you. Please just reply directly to me and
> not to the SBI mailing list.
>
>
>
> Thanks!
>
>
>
> Mark
>
>
>
> *Dr. MARK A.** MILLS*
> Professor of Mathematics | Central College
> 812 University Street | Campus Box 06 | Pella, Iowa 50219
>
--
Nathan Tintle, Ph.D.
Associate Professor of Statistics and Dept. Chair
Director for Research and Scholarship
Dordt College
Sioux Center, IA 51250
nathan.tintle(a)dordt.edu
Phone: (712) 722-6264
Office: SB1612
Instructors,
As you may know, we recently received NSF funding (DUE-1323210) to
facilitate assessment of (algebra-based) introductory statistics courses,
with a focus on gaining a better understanding of potential differences in
student learning between “traditional” and simulation/randomization-based
introductory statistics courses. As such, we are asking you to consider
having your students participate in the assessment project *regardless of
how much (if any) simulation- and randomization-based inference methods you
use in your course*. As a thank you for your participation, we are happy to
offer a $100 stipend and a customized report on your students’ performance
in your class. If you are interested in participating, please fill out this
short survey, as soon as possible, but early enough to allow time to set up
individualized links for your class before your term starts:
https://www.surveymonkey.com/s/9SYS8H3. Instructors are signing up now for
classes which begin after January 1.
Some brief details follow, with answers to some commonly asked questions
here <http://homepages.dordt.edu/ntintle/faqs.pdf>.
1. Students in your introductory statistics course (undergraduate or
high school level) will take a pre-test (preferably before the course
starts, but no later than the first week of classes). The “test” contains
multiple choice questions that assess conceptual understanding and student
attitudes toward statistics. Most students take approximately 45 minutes
to complete the test. The test is administered completely online (we will
provide you the link). Students can complete the test either inside or
outside of class. See the FAQ for more information on encouraging student
participation and IRB including opt-out options. After the closing date
you specify, we will send you the student names and individual performance
data. At the end of your course, students will take a single multiple
choice, online post-test about attitudes and concepts (or, for the
post-test only, separated concepts and attitudes tests). Again, we’ll
provide you the individualized link.
2. Assuming your sections attain at least 75% participation rates, and
you fill out a brief (<30 minute) survey about your course (e.g., size,
pedagogy, classroom technology, etc.) at the conclusion of the course, you
will receive the stipend and customized report.
We anticipate publishing a series of articles on the data gathered as part
of this project. Neither student-nor instructor-level information will be
reported individually (only in aggregate) in these articles. If you are
interested in participating but have questions about your institution’s
IRB, please contact us as well.
Again, if you are interested in participating, please fill out this short
survey, as soon as possible, but early enough to allow time to set up links
for your class before your semester starts:
https://www.surveymonkey.com/s/9SYS8H3 . After we receive your survey
responses, you will be contacted directly by us with more
information/details.
Please direct additional questions to either Cindy Nederhoff (assessment
administrator: cindy.nederhoff(a)dordt.edu) or Nathan Tintle (project
director: nathan.tintle(a)dordt.edu).
Thanks for considering this,
Nathan Tintle (on behalf of the PIs: Nathan Tintle, Beth Chance, Dennis
Pearl, Soma Roy and Todd Swanson)
--
Nathan Tintle, Ph.D.
Associate Professor of Statistics and Dept. Chair
Director for Research and Scholarship
Dordt College
Sioux Center, IA 51250
nathan.tintle(a)dordt.edu
Phone: (712) 722-6264
Office: SB1612
SBI users,
We have archived SBI related webinar and e-conference presentations on the
blog here: https://www.causeweb.org/sbi/?p=947
This is a great way to hear more about the use of SBI.
Have a great weekend!
Nathan
--
Nathan Tintle, Ph.D.
Associate Professor of Statistics and Dept. Chair
Director for Research and Scholarship
Dordt College
Sioux Center, IA 51250
nathan.tintle(a)dordt.edu
Phone: (712) 722-6264
Office: SB1612
SBI colleagues,
I wanted to draw your attention to three things.
1. On behalf of my co-authors and Wiley, we are happy to announce that the
1e of "Introductions to Statistical Investigations" is now available. See
the email below for an official announcement from Wiley.
2. We still have room in our January workshop (Robin Lock, Todd Swanson and
Nathan Tintle presenting, Tuesday prior to JMM in Seattle). Please share
this workshop with your colleagues who may be 'on the fence' about SBI or
sign-up yourself. While there is no official deadline, sign-ups for this
free workshop are preferred by Nov 1 if possible.
https://www.causeweb.org/workshop/jmm16_investigation/
3. If you have any ideas for the SBI blog (www.causeweb.org/sbi) please
email me so that we can consider your idea for future writing pieces. If
you haven't visited in a while, go take a look at some of the great posts
on a variety of topics including why people made the switch, using
technology, the hardest thing about getting started and more.
Thanks, Nathan
------------------------------ 1e announcement email
On behalf of Wiley and the entire author team, we are pleased to announce
that *Introduction to Statistical Investigations* 1st Edition (ISI) by
Tintle, Chance, Cobb, Rossman, Roy, Swanson, and VanderStoep, is now
available and accompanied by WileyPLUS Learning Space. WilyePLUS Learning
Space is a brand-new technology that will enhance your students’ learning
experience. Concept Check questions will be included with each learning
objective and our 200 author-created videos, as well as our 1,130-plus
exercises, will now be assignable. With WileyPLUS Learning Space, your
students will be able to collaborate directly with group projects. So
whether you teach in a face-to-face, hybrid, or online model, you can still
assign the Investigations and Explorations as group projects.
For spring classes, you will have the ability to adopt ISI, with WileyPLUS
Learning Space. If you would like AN INSTRUCTORS REVIEW COPY OR to find
out more about WileyPLUS Learning Space, please click here to contact your
representative.
<http://professor.wiley.com/CGI-BIN/LANSAWEB?PROCFUN+PROF1+PRFFN15>
Take a look inside.
<https://www.dropbox.com/s/q70p9kjefofeck3/Tintle%201e%20pgs%2022-23.pdf?dl=0>
Best regards,
John LaVacca III
*John LaVacca III*
Marketing Manager, Mathematics & Statistics
Learning Solutions
--
Nathan Tintle, Ph.D.
Associate Professor of Statistics and Dept. Chair
Director for Research and Scholarship
Dordt College
Sioux Center, IA 51250
nathan.tintle(a)dordt.edu
Phone: (712) 722-6264
Office: SB1612
Announcing SpIntro-Stats Web Applets
/Jim Robison-Cox and Allison Theobold, Montana State University/
/October 13, 2015/
Software applications are an essential component of any simulation based
inference (SBI) statistics course. At least two suites of web applets
for introductory statistics currently exist
(http://lock5stat.com/Statkey) and (http://rossmanchance.com/applets)
and each allows public access. With the increasing interest in the
simulation approach, we think it time for an open source software
project focusing on SBI for introductory statistics which uses the
interactive capabilities of web applications. We have begun a framework
for such apps based on the Shiny web interface to R. We offer these to
the SBI community for comment, feature requests, and collaboration.
The repository for the code is
https://github.com/MTstateIntroStats/SpIntro-Stats.git
Running Spintro-Stats
There are several ways to run the web apps:
1.
To check features without installing you can view it in this web page:
http://shiny.math.montana.edu/jimrc/IntroStatShinyApps/
Please do not send students to this site, as we can’t handle hordes
of users.
2. To test on an individual computer running R (free and open source
software):
* Install packages: ggplot2, gridExtra, shiny, shinythemes and their
dependencies.
*
Enter these commands into the R console:
|library(shiny) runGitHub("MTstateIntroStats/SpIntro-Stats") ## Be
patient. A browser window will appear with the app inside. ## Use
control-C in the R console to exit. |
3.
If you want to use these apps in a classroom setting, you’ll need
some IT expertise. Specifically, a “Shiny server” which is free for
academic purposes, from https://www.rstudio.com/products/shiny/
Be sure git is installed, and clone the code as in:
|git clone https://github.com/MTstateIntroStats/SpIntro-Stats.git |
Then you will have your own site to serve your own students.
4.
To get your own copy of the code, you need “git” installed and a
github account.
Clone the repository
https://github.com/MTstateIntroStats/SpIntro-Stats.git.
See a git tutorial for more instructions: https://help.github.com/
Basic layout
To run the apps, the user has to choose the /Type/ of data to analyze:
* One Categorical,
* One Quantitative,
* Two Categorical,
* Two Quantitative, or
* One of Each – One Categorical & One Quantitative.
The first option under each variable type is [Enter/Describe Data] which
must be run before one can proceed to the next choices under each
Variable Type: [Test] or [Estimate].
*
Testing is performed by sampling from the null model (One
Categorical) or via permutation test (Two Categorical or
One-of-Each) or by bootstrap (Two Quantitative). The “One
Quantitative” test of a single mean slides the data over to have the
hypothesized mean, then resamples from the shifted distribution.
*
Confidence intervals are all based on bootstrap resamples. Plots
show percentile intervals, but summary statistics are displayed for
those who want to build /Estimate ± Multiplier * SE/ intervals (We
do both).
The apps also include
* a demo of confidence interval coverage (under “One Categ”)
* an animated demo of the bootstrap process (under “One Quant”),
* demos of the effect of randomization on a “lurking” variable (under
“One Categ” and “One Quant”)
* “spinner” and “mixer” (balls drawn from a box) apps for generating
random categorical variates (under “One Categ”), and
* z and t probability and quantile “look up” pages.
Contributing
To correct errors or suggest features please use the google group:
https://groups.google.com/forum/#!forum/spintro-stats
Suggestions are welcome, but given that this is a volunteer project, it
might take some time to implement feature requests.
If you would like to help with the project, the level of R programming
needed is moderate. However, some knowledge of shiny, of html, and
possibly of JavaScript is necessary.
Related Materials
We have also developed activities which utilize the web apps for a one
semester statistics intro course. See our other github repository:
https://github.com/MTstateIntroStats/IntroStatActivities and
https://groups.google.com/forum/#!forum/mt-state-intro-stat-activities
Contact: Jim Robison-Cox: jimrc(a)math.montana.edu
Hello Simulation-Based Inference (SBI) folks,
Nathan Tintle, Robin Lock, and Todd Swanson will be leading a (free)
one-day workshop in Seattle on Tuesday, January 5, 2016, immediately
prior to the Joint Mathematics Meetings. We would greatly appreciate
your helping to spread the word about this workshop to your colleagues
who might be interested in learning about teaching introductory courses
with an SBI approach. Please point them to the following link for more
information and a registration form:
https://www.causeweb.org/workshop/jmm16_investigation/
Thanks, and have a good weekend,
Allan
--
Allan J. Rossman
Professor and Chair
Statistics Department
Cal Poly
San Luis Obispo, CA 93407
arossman(a)calpoly.edu
http://statweb.calpoly.edu/arossman/
"There's no limit to what you can accomplish in life if you don't care who gets the credit."
Folks interested in simulation-based inference,
1. We are offering a free webinar on Tuesday involving professors who
recently made the switch to simulation-based inference and their
reflections on making the switch. Sign-up here:
https://www.causeweb.org/webinar/teaching/2015-09/
2. Remember that there are many excellent articles related to teaching with
simulation-based inference on the blog here: www.causeweb.org/sbi
3. We are in the process of developing a series of additional articles for
the blog. If you have suggestions on what you'd like to see on the blog
(questions answered, more on a particular topic), please email me those
ideas directly.
Thanks!
--
Nathan Tintle, Ph.D.
Associate Professor of Statistics and Dept. Chair
Director for Research and Scholarship
Dordt College
Sioux Center, IA 51250
nathan.tintle(a)dordt.edu
Phone: (712) 722-6264
Office: SB1612
Hi folks,
CAUSE is offering two free workshops prior to the Joint Mathematics
Meetings in Seattle in January, one of which addresses the topic of
simulation-based inference:
Monday, January 4, 2016: Bringing Passion to Your Introductory
Statistics Classroom: A Supportive, Multi-Disciplinary, Project-Based
Approach, presented by Lisa Dierker and Dennis Pearl
Tuesday, January 5, 2016: Teaching the Statistical Investigation Process
with Randomization-Based Inference, presented by Nathan Tintle, Todd
Swanson, and Robin Lock
More information and registration forms are available at:
https://www.causeweb.org/workshop/
Best wishes,
Allan Rossman
--
Allan J. Rossman
Professor and Chair
Statistics Department
Cal Poly
San Luis Obispo, CA 93407
arossman(a)calpoly.edu
http://statweb.calpoly.edu/arossman/