Demonstrating an Interactive Song for Learning Introductory Statistics

Presented by
Dennis Pearl, PhD, Pennsylvania State University (PA), and John Weber, PhD, Georgia State University (GA)

This video poster shares a walk-through talk-aloud demo of an interactive song, starting with the pre-song prompts the user responds to that yield inserted words in the completed song, roughly in the style of the Mad Libs word template game. Some of the student inputs involve making conceptual connections while others involve providing context or examples. With the support of our NSF-funded Project SMILES (Student Made Interactive Learning with Educational Songs for introductory statistics, as mentioned in the VOICES presentation by our fellow SMILES PI Larry Lesser), some two dozen interactive songs and a computer auto-grading interface were created and the effectiveness of the innovation for reducing college students' statistics anxiety and increasing their learning is currently being assessed with randomized experiments. We will touch on challenges and tradeoffs we have had to negotiate (with the help of Penn State's Bob Carey and University of Texas at El Paso's Dominic Dousa and Steve Haddad) in terms of aesthetics, pedagogy, and technology. We will also provide data from our spring and summer 2017 pilot tests on the student reactions to the innovation and we welcome your constructive feedback as well!

Rick Heineman

Fascinating idea! I am looking forward to the randomized test - it seems like a very effective way to find the impact of the songs themselves if everyone will already be answering the questions the same.

Fri, 09/22/2017 - 16:52 Permalink
Dennis Pearl

In reply to by Rick Heineman

Thanks for the kind words Rick. Indeed ... moving from in-class satisfaction studies to out-of-class efficacy studies sets a much higher bar. I'm keeping my fingers crossed that we'll be reporting good news at next year's VOICES conference.

Tue, 09/26/2017 - 13:07 Permalink

Thanks for this very useful look at this project, guys!  When I originally heard about the project, I imagined that students would be looking at the lyrics, with occasional blank spaces, and inserting key phrases in response to prompts, thus contributing to the writing of the song by taking into account meter, tone, etc. in addition to content.  The way the project is actually working, though, the students are blinded to the actual song until they have entered their inputs. Was there ever any consideration given to doing it "my way"?  It may be a minor point, but students *might* be more invested in the songs (and thus more inclined to use them, etc.) if they felt that they had contributed to them in a sort of songwriter role (rather than as people whose answers to questions happen to get inserted into the songs). 

Sun, 09/24/2017 - 18:10 Permalink
Dennis Pearl

In reply to by crowther

That's an interesting suggestion Greg (not one we ever focused on originally since we were thinking win terms of a "MadLibs" approach at the time). I agree it would probably result in students being more vested in the songs. One of the nice features of the way we went is that the prompts themselves provide a pre-intervention assessment of student knowledge on the concepts being asked about.

Tue, 09/26/2017 - 13:11 Permalink
Lynda Blodgett

I like the way your study was designed. Congratulations on the improved student performance.

Tue, 09/26/2017 - 01:51 Permalink
Monty Harper

Hi! I contributed some songs to this project, and it was great to actually see how students' inputs get inserted. I'm glad that even when students didn't like the music so much, the songs still eased their anxiety and helped them learn. I wonder whether their affinity toward the music correlates with how much it helps, or if the two are unrelated. It would be fun if students could dial up a musical style of their own choosing!

Tue, 09/26/2017 - 10:48 Permalink
Dennis Pearl

Thanks for the kind words Monty - and thanks for your extremely valuable contributions to the project as one of the key members of our artists collaborative. Here's some data relevant to the question you asked about student perceptions of the value of the interactive songs versus their affinity for the music:
1) for students who did not care for the music (tempo/genre/whatever)
74% felt it relieved anxiety; 67% found it helpful for learning; and 93% found it relevant to course material
2) for students who did like the music (tempo/genre/whatever)
82% felt it relieved anxiety; 84% found it helpful for learning; and 92% found it relevant to course material
So some difference in the groups but still fairly good news even from those who preferred a different style.

Tue, 09/26/2017 - 13:24 Permalink
Tom Heinzen

Nice, clever approach that looks promising. We've been applying principles of game design in our lab. You are focusing on STEM and Stats while we focus on college completion rates, but I suspect we are informed by similar ideas. Nice work. Thanks.

Tue, 11/21/2017 - 15:03 Permalink