Dot plot touchup

[flannery-denny]

Please make a new branch for this work.

There are some instance where squished dot plots stack dots in such a way that they mislead about peaks.
We need to make sure that the dot plot images we use support student thinking rather than frustrate.
To that end, we'll need to update images on the following pages (it would be ok to use fewer images if that feels like less work!)

• Comparing Dot Plots and Histograms
• Matching Dot Plots and Histograms - images for 3 and 4 having annoying stray high dots
• Matching Dot Plots and box plots - images for 3, 4 and 6 have distracting stray high dot plots

Separately,

• Should Our Class' Name Length Data refer to the Live Pyret Survey?
• Variability of Dot Plots needs a little bit of tlc
• the left column ins too narrow for the parentheses of the question numbers to show
• question 3 has two of the same dot plots
• Variability of Animals Weights - Make Your Predictions - while the ranges of cat and rabbit weights in our datasets is different, I don't think that corresponds to intuition people who are familiar with rabbits and cats would have. I suggest:
• either slim this down to just look at tarantulas and rabbits
• replace cats with dogs

Relatedly,

• Histograms and Variability
• The solution for "Explain how you arrived at your answer" requires the histograms to have a scale. Either: update the images or update the solution.
• The solution on this page states that we should ask ourselves, "how much the data points deviate from the mean". We don't seem to offer up this advice anywhere else! where does it belong?
• Histograms and Measures of Center
• Question 1: In order to be able to find mode(s) from a histogram the bin size has to be one and the values have to all be whole numbers. While we don't usually make histograms of this variety, ES and FD can live it with as an exception for this question. Simplest fix: update the axes on the images and the question text to say that there are only whole numbers.
• Question 4: This feels very forced "Times were rounded to the nearest 5 minutes." If you don't think the answer can be arrived at without it, resort to the same solution as for question 1.
• confirm similar fixes aren't needed for histograms and measures of center 2.

Should we add a dot plot images to the box plots notes page to get at the following statement?

• Because the widest section of the box plot is between 235 and 310 pounds, we understand that the weights of the heaviest 25% fall across a wider span than the others.
  • 310 may be an outlier
  • the weights of the players weighing between 235 pounds 310 pounds could be evenly distributed across the range
  • or all of the players weighing over 235 pounds may weigh around 310 pounds.