Welcome to the StatPREP video. StatPREP is an NSF-sponsored program to help college instructors of introductory statistics incorporate data-centric methods into their courses, and is in the third year of a five year grant. Are you looking for ways to get started? Do you require students to use technology to calculate statistics? We provide what's needed to add computation to your course, using methods as simple as mouse-based apps all the way to tutorials in the R language. We’ve developed lots of online materials that you can use in your courses, and we offer help to individual instructors who are trying to get started. Would you like to join a community of educators working on these topics.  The StatPREP team wants to hear about ways that you are incorporating data science principles into your elementary statistics courses.

I wish my introductory statistics class had used StatPREP, what a great idea! I was wondering how you are measuring the impact of your project on college faculty and students. Have you found that the professional development for mathematics faculty has changed their teaching practices? 

I'm intrigued by this work. I've been part of the statistics education community at K-12 for some time and there's some important things we've learned about how to help students (and teachers) grapple with and understand key ideas about types of variability, characteristics of distributional shape and aggregate measures, among other things. Several years ago I had an NSF-funded project looking at how working with "big" data sets differed from working with smaller data sets, and I think there continues to be a tension between the groundedness of data that students collect themselves to answer their own questions, and the power that comes from working with larger, already cleaned and curated data. I'm curious what you're finding, especially with your community college students.

For me personally, my community college students are finding that they can answer more questions with larger, more robust, data. During my class, they spend time with a dataset and I am finding them asking more questions that the data can help them answer. I am finding that the more students are exposed to real data they are exploring more questions that they can ask. In my class they find data that will answer a question and they were finding much larger datasets then they did in the past.

We will need to ask our participants if they are having similar experience with their students. 

That's really cool work! I have had positive experiences from working with real-world large data sets with high school students. Students are awed when you say "let's explore this dataset with 2 million entries, 60 variables, shall we?" Then they realize the power of data exploration if you have the right tool in hands. They ask more and more questions that the dataset can help them to answer and - which I find extremely positive - some other questions that will lead them to look for MORE data. When you reach the point where students are actively gathering more data, I think we got them! They are now engaged! ;)

Very very nice work, indeed!

Kathryn,

Thanks for sharing this work! Quick question: I don't have any experience with the R language (although I am an old-school programmer). Do students in an introductory statistics course find it difficult to learn, or is it straightforward?

Kathryn,

Thanks for your video and discussion.  In your class, are students able to choose databases that provide data about topics that particularly interest them?  Do they locate data themselves, or do you provide specific clean, curated databases?  We are interested in providing data science experiences for high school students, and would like to enable them to have as much voice and choice as possible.  But that introduces all sorts of complications.  So I'm interested in how you approach it.

Marion

This looks fantastic - very exciting! These are definitely critical skills for faculty/instructors, and of course for their students in turn. Since there is so much variation in what intro stats courses cover in higher ed, I'm wondering if that's been a challenge you've encountered for making your PD & curricular materials relevant for a broad audience of faculty/instructors, and if so, what you've done to address it?

Thanks for the comment. There is variation in intro stats courses, but there are also 5 to 6 major topics taught in every statistics course. Those topics are: what is data and graphing data, descriptive statistics, the normal curve, hypothesis testing, confidence intervals, and linear regression. The Little Apps that are created for the project focus on those topics. We have found that showing how to infuse data into these topics helps to bring statistics classes into a data centric mind set. We have also created a Little App to show faculty more modern methods of statistics. We feel that we have been able to show introductory statistics faculty/instructors materials that will allow them to bring data into a statistics course.

Thank you for sharing your project and work!  Very important and relevant, and I really appreciate your comments about students actually asking MORE questions and having to explore more data by using large data sets!  I am wondering if you could tell us a bit about the scope of your project?  How many Statistics faculty have been engaged?  Have they all been local?  How are you thinking about scaling?  Thanks!

Thank you for your questions. The project picks different metropolitan areas around the country to host workshop. So far we have had workshops in Minneapolis, Los Angeles, Hartford, and Seattle. The Minneapolis and Los Angeles has had two years of summer workshops, with about 30 people attending the first year and then 39 people attending the second year. Some of the second year participants were repeats from the first year. Seattle and Hartford have had there first year workshop with about 30 people attending each. This summer they will have their second year with about 30 people attending this year, with some repeats. This summer we are also having 1st year workshops in Fort Worth and Washington DC area. There are 30 participants at Fort Worth and 45 at the Washington DC area workshop. Next summer both Fort Worth and DC will have their second year workshops. New York City and Fort Meyer will have their first year workshops at that time too. The following summer they will have their second year workshops. That will be the end of the grant.  

The idea is that we create local hubs of faculty in each of the workshop areas for them to share what they are doing. The issue we are having is creating this local hub where they keep in contact. Keeping the participants from the first two workshop locations engaged is a challenge for the grant, so any advice on keeping them engaged would be appreciated. The idea is that they will then engage other faculty in their area which will help to scale it. Any advice on scaling would also be appreciated.

Hi Kathryn, as you describe having repeat workshop participants, I'm curious what kind of feedback you've gotten from them after they've used the materials for a year?

Also, the scaling and sustainability pieces are always so tough. One thing that comes to mind is whether there are existing local networks that you could tap into, rather than trying to build new networks. And whether there are some key, trusted leaders trusted within those networks to potentially connect with (although by your description, I think you may already be using this approach.) You may be interested in reading this to think about spreading use and scaling: https://press.princeton.edu/titles/11279.html

WE haven't collected that data yet, but we need to. I will ask our evaluators when they will be asking that information. They have collected information on what participants are planning to do, and most are planning to incorporate more data into their classes. We have retooled the workshops based on comments from participants, and I am interested to see how this year's workshops are received.. 

We have what we call hub leaders, but we need to give them some direction of the type of connections they need to be making. We just have to create a communication plan that will help them see how to maintain the connections made at the workshops.

The book you suggested sounds interesting. I will suggest it to the leadership team. Thanks.

I love this idea! I work with faculty from many disciplines, and one thing they keep bringing up is the need for students to understand how to work with large data sets. Your project is aimed at integrating those data sets into stats classes. Would it be feasible to integrate the materials you've created into general STEM classes? That is, could some of the materials be used in isolation rather than as part of a broader statistics class? 

Great question. The material we created is to teach students statistical concepts using data, but I would think you could use the material in a class that is a STEM class. The materials use visualization of the data to understand statistics, and I would think in STEM classes visualizing the data would be useful. visit statprep.org, and then go into Resources. In Resources there are Little Apps that do not require anyone to use R or create the graphs. The graphs are there and then you can pick different data frames. From there you can pick different explanatory and response variables. Let me know if you have any other question after you look at the Little Apps. There are also R tutorials, that help the students learn R commands, but also can be used to look at large datasets to look at trends of the data.

I was talking to some folks recently about algebra as a gatekeeper. I am happy that AMATYC is working on this. Good luck. I'll check the resources you mentioned and pass them on. 

Thanks. Unfortunately, algebra is still the gatekeeper in some states. In my state, College Algebra or Liberal Arts Math is still the prerequisite to statistics. There is work to make the Liberal Arts Math class not have as high of a level of algebra though. StatPREP doesn't actually say what the prerequisite of statistics should be, but hopefully when more institutions see that statistics is not dependent on algebra there will be more of a discussion about what the prerequisite should be. Let me know if you have any questions when you look at the resources, and also I would love to hear your thoughts on algebra as a gatekeeper.

What a great project! I especially like that you are creating local faculty hubs through your workshops. You mentioned above that you're interested in ideas about how/where to keep them in contact. I would be very interested in hearing what you've tried already. Have you done any coordinating with the institutions the faculty represent, to see if they might commit some very small amount of support (a place for convening, assistance organizing)? And/Or could some of the participants be developed as "hub leaders" (with some additional incentive/support) to be responsible for convening a group?  

Andrea, thank you for you questions and suggestions. Each faculty hub already has a hub leader that is suppose to keep them in contact. I think the problem that we have is that we haven't given the hubs a great deal of guidance in terms of how often to contact and what they should be contacting with. While answering your questions, I was just thinking that maybe we should give the hub leaders a timeline and a script of what to email so that they can keep their hubs active. At least the script and timeline may keep the hubs engaged at the start. Thank you for helping me think about this problem.