New AP Stats Teacher Moves Using Desmos

Last spring, the awesome folks at Desmos released a slew of slick (but easy-to-use) statistics features. Here is a brief video I made which walks through a few of the new features. With a new academic year beginning, I’m looking forward to changing some of my classroom moves in AP Stats to leverage the new features and build understanding. Here are 3 moves I’m planning to try this year:

ASSESSING NORMALITY (Here is a previous post on this topic)

Pop quiz! Below you see 6 boxplots. Each boxplot represents a random sample of size 20, each drawn from a large population. Which of the underlying populations have an approximately normal shape? Take a moment to think how you…and your students…might answer…

6 random samples (n=20) from “large” populations

Have your answers ready? Here comes the reveal…..

Not only do each of the samples above come from normal populations, they each come from the same theoretical population! This year in class I plan to walk students through how to build their own random sampler on Desmos, which takes only a few intuitive commands. When the “random” command is used, we now get a re-randomize” button which allows students to cycle through many random samples and assess the shapes. You can toy with my graph here.

Often students look for strict symmetry or place too much stock in different-sized tails. This is a great opportunity to have students explore and understand the variability in sampling. Teach your students to widen their nets when trying to assess normality and remember – our job is usually not to “prove” normality; instead, these samples show that the assumption of population normality is often safe and reasonable, especially with small samples.

LINEAR TRANSFORMATIONS OF DATA

Analyzing univariate data using Desmos is now quite easy. Let your students build and explore their own data sets. Data can be either typed in as a list or imported from a spreadsheet using copy/paste. The command “Stats” provides the 5-number summary, and commands for mean and standard deviation are also available. You can play around with my dataset here.

Next, I want my students to consider transformations to the data set. In my example I have provided a list of test scores and summary statistics are provided. Let’s think about a “what if”. In the next lines I provide 2 boxplot commands, but I have intentionally ruined the command by placing an apostrophe before the command (thanks Christopher Danielson for this powerful move!). What will happen if every student is given 5 “bonus” points? What if I feel generous and add 10% to everyone’s grade?

What will happen when I remove those apostrophes? Think about the center, shape and spread of the resulting boxplots? How will these new boxplots be similar to and different from the original?

Compute new summary statistics. Which stats change…by now much…and what stays the same? Why? I’m looking forward to having students build their own linear transformation graphs, investigating and summarizing their findings! Here is a graph you can use with your classes to explore these linear transformations with sliders.

COMBINATIONS OF DISTRIBUTIONS

An important topic later in AP Stats – what happens when we combine distributions by adding or subtracting? Often I will use SAT scores as a context to introduce this topic because there are two sections (verbal and math) and a built-in need to add them – What are the total scores? On which section do students tend to do “better”…and by how much? To build a Desmos interactive here, I start with a theoretical normal distribution with mean 500 and standard deviation 100 to represent both mean and verbal score distributions. Next, taking 2 random samples of size 1000 and building commands to add and subtract them allows us to look at distributions of sums and differences and compare their center, shape and spread.

The most important take-away for students here should be that distributions of sums and differences have similar variability. This is a tricky, yet vital, idea for students as they begin to think about hypothesis tests for 2 samples. You can use my graph, or build your own. Note – in my graph the slider is used to generate repeated random samples.

Advertisements

Techstravaganza Resources

Desmos and Statistics! Resources from the AP Statistics Reading “Techstravaganza” – June 2019 in Kansas City, Bob Lochel and Leigh Nataro

PCTM 2019

I am proud to be the program chair for the 2019 PCTM Conference in August, with my good friend Sue Negro.

The preliminary lineup of speakers appears in the file below. With Keynotes by Dan Meyer and Robert Berry, a Desmos pre-conference the day before, and our first ever trivia night, it’s going to be a great 2 days in Harrisburg!

Register today and get all the details:
https://www.eventbrite.com/e/pctm-annual-conference-registration-52503521446

Golfing with Linear Equations

For the first time in many years, I am teaching a College Prep Algebra 1 class with a fantastic group of 9th graders. Nearing the end of our linear functions unit, my colleagues and I discussed a desire to have some sort of culminating activity. And while I have used drawing projects often in some courses, in algebra 1 such tasks have often left me feeling unfulfilled. Too many horizontal and vertical lines for my liking I suppose.

I recalled reading about a potential golf-related task on twitter. To be honest, I don’t recall whose exact post provided the inspiration here (note – I am thinking it was Robert Kaplinsky or John Stevens, but I may be wrong. If anyone locates a source, I’ll edit this and provide ample credit), but it felt like a game-related task could provide by the strategy and fun elements which tend to be missed by drawing tasks.

HOW THE GOLF CHALLENGE WORKS:

The goal – write equations of lines which connect the “tee” to the “hole”. Use domain and/or range restrictions to connect your “shots”. Try to reach each hole in a minimal number of shots. Leaving the course (the green area) or hitting “water” are forbidden. All vertical or horizontal likes incur a one-stroke penalty.

On the day before the task, the class worked through a practice hole. Besides understanding the math task, there are also a few Desmos items for students to understand:

  • Syntax for domain / range restrictions
  • Placing items into folders
  • Turning folders on /off

practice hole

golfFor the actual task, a shared a link to a Desmos file with 5 golf holes. I tried to build tasks which increased in their difficulty. In practice, the task took an entire class period (75 minutes), and students worked in pairs to discuss, plan, and complete the holes. All students then uploaded their graphs to Canvas for my review, and filled out a “scorecard” which included “par” for each hole.  It became quite competitive and fun!

hole1hol4

hole5

In the end, there is not too much I would change here. Perhaps add some more complex holes. I’d also like to provide opportunity for students to design and share their own golf holes, and study the “engine” which built mine.  I hope your class has fun with it! Please share your suggestions, questions and adaptations.

What’s Going On in This Graph

Today the New York Times Learning Network dropped the first “What’s Going On in This Graph?” (WGOITG) of the new school year. This feature started last year as a monthly piece, but now expands to a weekly release. In WGOITG, an infographic from a previous NYT article is shown with the title, and perhaps some other salient details, stripped away – like this week’s graph…

12GraphLN4-jumbo

Challenge your students to list some things they notice and wonder about the graph, and visit the NYT August post to discover how teachers use WGOITG in their classrooms. Here are some ideas I have used before with my 9th graders:

  • Have groups work in pairs to write a title and lede (brief introduction) to accompany the graph.
  • Ask tables to develop a short list of bullet points facts which are supported by the graph, and share out on note cards.
  • Have students consider how color, sizing, scaling are used in effective ways to support the story (note how the size of the arrows play a role in the graph shown here). This is a wonderful opportunity to think of statistics beyond traditional graphs and measures.

Invite your students to join in the moderated conversation, which drops on Thursday. Have your own favorite way to use WGOITG? Share it in the comments!

 

 

TMC Desmos Day – Stats Session

Today I am in Cleveland for 2018 Twitter Math Camp! It’s the Desmos Pre-Conference Day, and I am facilitating a session on using Desmos in Statistics classes. Below are many of the links and resources I plan to use – even if you are not in Cleveland with us, feel free to borrow from these resources.

Baseball Data Set

Comparing Data Sets and Summary Statistics

Regression Facts (Mean/mean point and slope)

Teaching the meaning of r-squared

“Release the Hounds” – my first attempt at random sampling

Participate as a student, Steal and Share

“Backpack Weights” – thinking about scatterplots (AP Stats)

Participate as a student, Steal and Share

Assembling the Model Solution

I use College Board released AP items often in my Statistics course. The problems are aligned to clearly-stated goals, and the solutions provide insight not only into the grading of AP questions but also allow students to study well-articulated explanations. You can visit the College Board Statistics website and explore. Jason Molesky’s website provides helpful guidance on using FRAPPY’s (Free-Response AP Practice…Yay!) as a formative assessment tool in AP Stats.

Each free-response solution begins with the “model solution” – the ideal explanation a student would provide for full question credit. It is not unusual for Statistics students to struggle with clear communication, and having students read and dissect the model solution can be helpful in strengthening statistical arguments. A few times this year, I have used the model solution as a formative assessment tool with an activity I call “Assembling the Model Solution”.

Here’s how it works – start with an AP Free-Response question with a narrative aspect. Today, I chose a problem which requires students to interpret a P-value, from 2009:

2009problem

The model solution contains a number of non-negotiable elements: a conditional probability, a reference to smple results, and the “extremeness” of results.

2009sol

Next, I took the model solution as broke it into small, strategic “bites”. At the same time, I added some parallel distractors and a junk phrase or two.

slips

Then, use a paper cutter and slice the Word document into phrase slices, and paper clip together. All students then received the problem and the slips of paper, with the challenge to assemble the model solution for part a of the problem.

 

The conversation were rich, and the teams mostly debated the salient aspects of the problem apprpriately. The biggest points of debate and incorrect solutions came from:

  • The difference between “sample” and “population” proportions.
  • The assumption of sameness in the treatments as the conditional aspect of P-value.

I have used this strategy a few times now, and continue to tweak how I provide the slips of paper. I’m also looking at digital options, but I like the social aspect of moving the slips of paper. The method is not ideal for everything in AP Stats, but there are a few areas in our curriculum where this fits in nicely:

  • Sampling and experimental design
  • Conclusions for inference procedures
  • Describing distribuitions.

You can download my file for this activity here.  Enjoy!

  • Credit to Jon Osters and the AP Stats glitterati who rightfully pointed out that my original post spelled “Yay!” incorrectly.