Data Science Data Visualization and Communication Questions and Answers 1 — Questions and Answers
Question 1: A data scientist needs to create a visualization to compare the distribution of salaries for data analysts, data scientists, and machine learning engineers. The visualization must clearly show the median, interquartile range (IQR), and potential outliers for each job title. Which type of chart is most suitable for this purpose?
- A series of pie charts, one for each job title.
- A stacked bar chart showing the salary ranges.
- A box plot with separate boxes for each job title. (Correct answer)
- A line chart plotting the average salary over the last five years.
Correct answer: A box plot with separate boxes for each job title.
A box plot is specifically designed to summarize the distribution of a numerical dataset. It visually represents the five-number summary: minimum, first quartile (Q1), median (Q2), third quartile (Q3), and maximum. This makes it ideal for comparing the median, interquartile range (Q3-Q1), and identifying outliers across different categories like job titles. [2, 6]
Question 2: According to Edward Tufte's principles of data visualization, what is the primary goal of maximizing the "data-ink ratio"?
- To increase the total number of colors and graphical elements to make the chart more visually appealing.
- To ensure that the chart is printed using the highest quality ink available.
- To devote the largest possible share of the graphic's ink (or pixels) to displaying non-redundant data information. (Correct answer)
- To make the chart's background and grid lines darker than the data elements.
Correct answer: To devote the largest possible share of the graphic's ink (or pixels) to displaying non-redundant data information.
The data-ink ratio, a concept by Edward Tufte, is the proportion of a graphic's total ink that is dedicated to displaying essential, non-redundant data information. [1, 3, 4] The goal is to remove extraneous elements ('chartjunk') like excessive gridlines, borders, and decorations to present the data as clearly and concisely as possible. [3, 5]
Question 3: A data analyst is preparing a presentation for an executive board to argue for increased investment in a marketing channel. Which of the following communication strategies is MOST effective for this audience?
- Presenting a complex statistical model with its full mathematical notation to prove the validity of the analysis.
- Showing a detailed table with raw, unaggregated data so the executives can perform their own analysis.
- Focusing on a clear and concise narrative that connects the marketing channel's performance directly to key business outcomes like revenue and customer acquisition cost. (Correct answer)
- Using highly technical jargon to demonstrate deep expertise in data analysis.
Correct answer: Focusing on a clear and concise narrative that connects the marketing channel's performance directly to key business outcomes like revenue and customer acquisition cost.
When presenting to executives, the primary goal is to support better decision-making by providing clear, actionable insights. [15, 17] This audience is most concerned with the 'so what' – the impact on business goals. A concise narrative that links data insights to strategic implications and business outcomes is far more effective than delving into technical details or presenting raw data. [17, 19]
Question 4: An analyst creates a bar chart to compare the monthly revenue of four different products. To better highlight the differences, the y-axis is set to start at $500,000 instead of $0, as all products generated over $520,000. What is the primary issue with this visualization choice?
- It makes the chart less colorful and engaging.
- It visually exaggerates the proportional differences between the products, potentially misleading the audience. (Correct answer)
- It violates the principle of maximizing the data-ink ratio by adding unnecessary complexity.
- It makes the chart impossible to read for anyone who is colorblind.
Correct answer: It visually exaggerates the proportional differences between the products, potentially misleading the audience.
Truncating the y-axis (i.e., not starting at zero) is a common way to create a misleading graph. [20, 25, 26] For bar charts, the length of the bar is a key pre-attentive attribute that viewers use to compare values. Starting the axis at a non-zero value distorts this visual comparison, making small differences appear much larger and more significant than they actually are. [25, 26]
Question 5: Which of the following visual elements is a pre-attentive attribute, meaning it can be processed by the human brain almost instantaneously without conscious effort?
- The exact value of a data point read from the y-axis.
- The average of all values in a bar chart.
- A single red dot in a field of grey dots. (Correct answer)
- The statistical significance (p-value) of a trend line.
Correct answer: A single red dot in a field of grey dots.
Pre-attentive attributes are visual properties that our brains process in milliseconds, before we pay conscious attention. [11, 23] Key attributes include color, form (length, width, shape), and spatial position. [8, 11] A distinct color, like a single red dot among grey ones, 'pops out' and is noticed immediately, making it a powerful tool for drawing attention to important data points. [8, 12]
Question 6: A researcher is investigating the potential relationship between two continuous variables: the number of hours a student studies per week and their final exam score. Which type of visualization is most appropriate for initially exploring this relationship?
- Pie Chart
- Bar Chart
- Histogram
- Scatter Plot (Correct answer)
Correct answer: Scatter Plot
A scatter plot is the standard and most effective chart for visualizing the relationship between two continuous (numerical) variables. [27, 29, 30] Each point on the plot represents a single observation (a student) plotted according to its values on the two axes (hours studied and exam score), allowing the researcher to identify patterns, trends, correlations, and outliers. [28, 30]
Data Science Data Visualization and Communication Questions and Answers 1
A data scientist needs to create a visualization to compare the distribution of salaries for data analysts, data scientists, and machine learning engineers.
The visualization must clearly show the median, interquartile range (IQR), and potential outliers for each job title.
Which type of chart is most suitable for this purpose?