DAO1704 Chap.2 Visual Choices and PivotTable Evidence
Visual Choices and PivotTable Evidence
Begin with variable type
A visual should make the intended comparison easier to see. Histograms and boxplots describe one numerical distribution; bar charts compare categorical frequencies or proportions; scatterplots examine two numerical variables; and side-by-side displays compare distributions across groups.
Choosing from the chart gallery before naming the variables reverses that logic.
Aggregation sits beneath every chart
A PivotTable assigns fields to Rows, Columns, Filters and Values. The Values field can count, sum or average, and those choices answer different questions. Total sales can rise because there are more orders even while average order value falls.
The chapter keeps volume and value separate so that a chart title cannot imply a conclusion the aggregation does not support.
Visual settings can create the pattern
Histogram width changes apparent shape, bar order changes emphasis and colour can make a small category look important.
The method tests nearby bin widths, cross-checks a histogram with quartiles and a boxplot, labels units and denominators, and closes with one limitation or alternative explanation that the present display cannot settle.
Titles should preserve the aggregation
A good title says whether the chart shows totals, counts, averages or shares and identifies the population and period.
This wording prevents a reader from translating a regional total into a claim about a typical customer or interpreting a scatterplot as a causal effect. Before exporting a figure, read the title, axes, legend and PivotTable value field as one argument. Then verify one bar or point against the underlying rows. That independent trace catches filter states, missing categories and units that the formatting can conceal.
What this chapter covers
- 01
Chart choice by variable type
- 02
Histogram width and binning sensitivity
- 03
Boxplots, quartiles and flagged outliers
- 04
PivotTable row, column and value roles
- 05
Counts, sums and averages as different questions
- 06
Scatterplots and third-variable encodings
Worked example · free
Separating transaction volume from transaction value
- 2Calculate each region's mean order value.
- 2Compare total sales and order counts.
- 2Write a conclusion that preserves both findings.
Key terms
- Histogram
- A display of numerical frequency across contiguous intervals.
- Bin Width
- The numerical span grouped into each histogram bar.
- Boxplot
- A compact display of quartiles, median, whiskers and flagged distant values.
- PivotTable
- An interactive summary that aggregates fields across selected categories.
- Aggregation
- The operation used to combine observations, such as count, sum or average.
- Bubble Chart
- A scatterplot that encodes a third numerical variable through point size.
Visual Choices and PivotTable Evidence FAQ
How should a histogram bin width be chosen?
Try several nearby widths and retain conclusions that persist. Too few bins can hide clusters or skew, while too many can make random fluctuation look structural. Cross-check with quartiles and a boxplot before describing the distribution.
Why can a PivotTable value field be misleading?
Excel may count a field when blanks or text prevent numeric aggregation. A polished chart then visualises record frequency rather than amount. Read the aggregation label, inspect the source type and reconcile a control total before interpreting the view.
When should total and average be shown together?
Show both when volume and value answer different parts of the decision. Keep their units visually separate, state the denominator behind the average and avoid a dual axis that encourages direct comparison of unrelated heights.
Exam move
Use one small dataset to make a histogram, boxplot, bar chart and scatterplot. For each display, write the question it answers and one question it cannot answer. Change the histogram width, swap a PivotTable from Sum to Count and reorder the categories. Record which verbal conclusions change.
This exercise builds sensitivity to chart settings faster than copying a finished dashboard because every change has a visible analytical consequence. Recreate one bar from the source rows and confirm its aggregation by hand. Then write two titles, one accurate and one that overstates causality, and explain the exact word that makes the second unsupported. This trains interpretation alongside chart construction.
Finish with a legend-free reading: cover the title and explain the axes, unit, aggregation and comparison aloud. Any element you cannot name needs a clearer label before the chart is ready.
Working through Visual Choices and PivotTable Evidence in DAO1704? Sia is AskSia’s AI Statistics tutor — ask any DAO1704 Visual Choices and PivotTable Evidence question and get a clear, step-by-step explanation grounded in how DAO1704 is taught and assessed. Read this chapter free, then take your hardest questions to Sia.
