Exploratory Analysis and an Honest Chart · seed 1 · A4, ink-friendly. The answer key prints on its own page for grown-ups.

Look first, then chart it honestly

Computing · Data & Databases · ages 21-22
Name ______________________   Date ____________

  1. You want to show how the temperature changed day by day across one month. Which chart fits best?

    • A pie chart of the days in the month
    • A line chart of temperature by day
    • A histogram of all the daily readings
    • A bar for the average temperature

  2. Which chart compares two sets of data?

    • Bar chart
    • Pie chart of one set
    • Histogram of one set

  3. Bar charts should start their axis at zero.

    Circle one:   True   False

  4. What do you make before computing correlation?

    • A correlation number, then a plot if needed
    • A scatterplot of the two variables
    • The mean and then stop

  5. You want to compare the distribution of commute times for cyclists and for drivers. Which display shows the comparison best?

    • A pie chart of the group sizes
    • One scatterplot of time against time
    • A single bar of the overall average
    • Side-by-side boxplots of the two groups

  6. Which of these is NOT a way to display categorical data?

    • Bar graph
    • Pie chart
    • Frequency table
    • Histogram

  7. A plot of practice hours against quiz score rises steeply, then levels off. Why look at this plot before reporting a correlation number?

    • Correlation only works when the dots slope downward
    • A plot is always more precise than any number
    • A single number can hide the curve the plot shows
    • The number would change if the dots were recoloured

  8. A frequency table of goals per game shows 0 goals in 2 games, 1 goal in 3 games, 2 goals in 4 games and 3 goals in 1 game. What is the mean number of goals per game? Round to one decimal place.

    Answer: ______________

  9. An outlier traces to a thermometer left in direct sun. What is the honest move?

    • Keep it to protect the average
    • Copy it into every row
    • Remove it and write down why

  10. A cut axis made sales look tripled. What did the original make readers believe?

    • That sales had fallen over the period
    • That the chart covered every month of data
    • That growth was huge when it was tiny
LightMySky · lightmysky.comW1-mt_Jpb78Z6Ytu-s1

Answer key

For grown-ups. Fold this page away before handing over the rest.

Look first, then chart it honestly W1-mt_Jpb78Z6Ytu-s1

  1. A line chart of temperature by day · Change over time needs the days in order on one axis, which a line chart keeps. The other charts lose the order or the change.
  2. Bar chart · Bars exist for listing differences and comparing similar sets.
  3. True · Starting at zero keeps small gaps looking small.
  4. A scatterplot of the two variables · A plot shows the shape that a bare number would hide.
  5. Side-by-side boxplots of the two groups · Side-by-side boxplots put the shape, centre, and spread of each group in one view, which a pie chart, an average, or a plot of a variable against itself cannot do.
  6. Histogram · A histogram bins a quantitative variable into number ranges. Categories like breeds or colours have no order or width, so they belong in bars, pies, or tables.
  7. A single number can hide the curve the plot shows · A correlation number measures a straight-line trend. It cannot say that the rise levels off, and the plot shows that at a glance.
  8. 1.4 · Multiply each value by its frequency, add the products, and divide by the total count: (0*2 + 1*3 + 2*4 + 3*1) / 10 = 1.4 goals per game.
  9. Remove it and write down why · A proven error may go, but only with the reason on record.
  10. That growth was huge when it was tiny · Stretched gaps sell a tiny rise as a boom, which the redraw exposes.
Worksheet · LightMySky