Data journalism: what you need to know

• A. Publishing opinion pieces backed by statistics
• B. Using data analysis and visualization to find, verify, and tell news stories ✓
• C. Replacing reporters with automated algorithms
• D. Writing only about financial markets and economics

Why: Data journalism uses data collection, analysis, and visualization to discover and communicate stories, combining traditional reporting skills with quantitative methods.

• A. A video interview recording
• B. A scanned handwritten letter
• C. A spreadsheet of government spending by year and department ✓
• D. A social media post with photos

Why: Structured data is organized into rows and columns with consistent fields, like a spreadsheet. Videos, scanned images, and social posts are unstructured.

• A. Older datasets are always more accurate
• B. To understand how the data was collected, its limitations, and whether it is reliable ✓
• C. Because the source must be cited in the headline
• D. To make the visualization look more professional

Why: Knowing a dataset's origin, collection methods, and gaps helps journalists avoid misleading conclusions and assess credibility before publishing.

• A. Deleting all data that shows unfavorable results
• B. Correcting errors, removing duplicates, and standardizing formats so the data can be analyzed ✓
• C. Encrypting the data before publication
• D. Converting every number into a percentage

Why: Data cleaning fixes inconsistencies, errors, and duplicates and standardizes formats. Deleting unfavorable results would be data manipulation, which is unethical.

• A. What chart type will look best on the homepage?
• B. What is the question or claim I want to test, and what data could answer it? ✓
• C. How can I get the largest possible audience quickly?
• D. Which software is most expensive?

Why: Defining a clear question and identifying relevant data first keeps the investigation focused and ensures the analysis can actually support or refute the claim.

• A. Does the dataset have a colorful visualization attached to it?
• B. Who collected the data, and what was their methodology? ✓
• C. How many rows does the dataset contain?
• D. Is the file in the most popular format available?

Why: Understanding who collected the data and how they did it reveals potential biases, gaps, and limitations. Size, format, and presentation matter far less than the provenance and method behind the data.

• A. The dataset is too recent to be reliable.
• B. Reported crime may not reflect actual crime, since many incidents go unreported. ✓
• C. Government data is always accurate, so there is no concern.
• D. The dataset should include private company records instead.

Why: Official crime statistics based on reports can undercount crimes that victims do not report, so the data may systematically understate or misrepresent true crime levels. Journalists should acknowledge this limitation.

• A. It was published most recently.
• B. It was collected directly by the original researcher or organization, rather than compiled from someone else's work. ✓
• C. It is the largest available dataset on the topic.
• D. It is free to access without a subscription.

Why: A primary source comes directly from the original collector, such as raw survey responses or official records. Secondary sources reuse or summarize data gathered by others, which adds a layer where errors or distortions can creep in.

• A. Metadata automatically corrects errors in the data.
• B. Documentation explains variable definitions, collection dates, and known limitations, which helps prevent misinterpretation. ✓
• C. Metadata is required by law for all published statistics.
• D. Documentation makes the file size smaller for faster downloads.

Why: Metadata and documentation tell you what each column means, how the data was gathered, when it was current, and what caveats apply. Without them, you risk misreading variables, such as assuming a field measures one thing when it measures another.

• A. Publish the spike immediately because it is newsworthy.
• B. Delete the anomalous year from the dataset to make the trend cleaner.
• C. Investigate whether the spike reflects a change in data collection methods or a data entry error before drawing conclusions. ✓
• D. Assume the spike is accurate because the source is official.

Why: Sudden unexplained changes often come from methodological shifts, reclassification, or errors rather than real events. Verifying the cause before reporting protects against publishing a misleading story.

• A. Duplicate rows make the file size smaller, which slows down software.
• B. Duplicate rows can inflate counts and totals, leading to misleading conclusions. ✓
• C. Duplicate rows are automatically removed by spreadsheet software, so checking is unnecessary.
• D. Duplicate rows only matter when working with text data rather than numbers.

Why: Duplicate records double-count observations, which can inflate sums, averages, and frequencies and lead journalists to report inaccurate findings. Spreadsheet programs do not automatically remove duplicates, so checking is necessary.

• A. Delete the entire column because it contains non-numeric values.
• B. Replace every 'unknown' with a zero so the column becomes fully numeric.
• C. Document how many values are missing and decide how to handle them before analysis. ✓
• D. Ignore the issue, since averages will automatically skip text values.

Why: Good practice is first to assess the extent of missing data and then choose an appropriate handling method, such as excluding those rows, flagging them, or using a justified imputation. Replacing unknowns with zero would distort averages.

• A. Making all text uppercase so it is easier to read on screen.
• B. Ensuring values like dates, currencies, and place names follow a consistent format throughout the dataset. ✓
• C. Converting every file into a PDF so it cannot be edited accidentally.
• D. Removing all columns that do not contain numbers.

Why: Standardization means consistent formatting, such as using one date format (YYYY-MM-DD) or consistent country names. Inconsistent formats can cause sorting errors, failed joins, and miscounted categories.

• A. Original files are legally required to be kept in every newsroom.
• B. Raw data cannot be opened once it has been modified.
• C. It allows the cleaning steps to be audited, verified, and repeated if errors are found. ✓
• D. Keeping the raw file prevents any cleaning from being necessary.

Why: Preserving the original data creates a transparent record of every change made. This supports reproducibility and lets others, or the journalist later, check whether a cleaning decision affected the findings.

• A. The numbers are in different languages and must be translated.
• B. The values are stored as text with inconsistent formatting, so they must be converted to a consistent numeric format. ✓
• C. The second value is incorrect and should be deleted.
• D. Population figures should never be stored as numbers in a dataset.

Why: Commas used as thousands separators can prevent software from recognizing a value as a number. Removing the separators and converting the text to numeric type ensures calculations work correctly.

• A. Pie chart
• B. Line chart ✓
• C. Scatter plot
• D. Treemap

Why: Line charts connect data points in time order, making trends and changes over time easy to see. Pie charts show parts of a whole at a single moment, and scatter plots and treemaps are better suited to other comparisons.

• A. To show the relationship between two continuous variables
• B. To compare values across distinct categories ✓
• C. To display geographic patterns on a map
• D. To show the composition of a dataset over time

Why: Bar charts are designed to compare quantities across discrete categories, such as countries, parties, or product types, allowing readers to quickly judge which is larger or smaller.

• A. Because most spreadsheet software requires it
• B. Because bar length encodes value, and a truncated axis can exaggerate differences ✓
• C. Because it makes the chart load faster on websites
• D. Because readers expect all charts to begin at zero

Why: With bars, readers compare lengths. If the axis does not start at zero, small differences can look dramatic, misleading the audience about the true magnitude of the difference.

• A. Microsoft Word
• B. Datawrapper ✓
• C. Adobe Acrobat
• D. Notepad

Why: Datawrapper is a widely used web-based tool among newsrooms for building clean, embeddable charts and maps without requiring coding skills. Word, Acrobat, and Notepad are not designed for interactive data visualization.

• A. It increases the file size for better print quality
• B. It allows readers to verify the data and understand how current it is ✓
• C. It is required to make the chart colorful
• D. It replaces the need for a headline

Why: Citing the source and date builds credibility and transparency, letting readers check the original data and judge whether the information is still current.

• A. To confuse readers with complex statistics before explaining them
• B. To grab the reader's attention and establish why the data matters to them ✓
• C. To show that the journalist has done extensive research
• D. To replace the need for any written context in the article

Why: A strong lede makes the stakes concrete for readers. A striking, relatable data point draws attention and signals why the story is relevant to their lives.

• A. Converting large numbers into relatable comparisons, such as showing how many people a figure affects in a familiar setting ✓
• B. Using only percentages instead of raw numbers
• C. Presenting data exclusively through charts without any text
• D. Limiting the story to one individual's experience with no data

Why: Human scale translates abstract figures into terms people can picture, like 'enough to fill a stadium' or 'one in every four households,' making the data easier to grasp and remember.

• A. It is optional and only adds unnecessary length to the story
• B. It allows readers to judge the context and reliability of the findings ✓
• C. It ensures the editor will approve the article faster
• D. It prevents other journalists from using the same data

Why: Context such as the time frame and source lets readers evaluate whether the findings are current, credible, and applicable, which builds trust in the reporting.

• A. Placing a figure alongside comparisons, historical trends, or benchmarks so its meaning is clear ✓
• B. Rounding the number to make it sound more impressive
• C. Removing all numbers from the text to keep it simple
• D. Hiding the number in a footnote so readers do not focus on it

Why: A figure like '5,000 cases' means little alone. Adding context, such as the change from last year or the rate per population, reveals whether it is alarming, normal, or improving.

• A. Emphasizing only the most dramatic figure and ignoring the rest of the dataset
• B. Showing the full picture, including limitations and uncertainty in the data ✓
• C. Using the largest possible y-axis scale to exaggerate differences
• D. Omitting the source so readers focus on the findings alone

Why: Acknowledging limitations, caveats, and uncertainty gives readers an honest picture. Cherry-picking or hiding weaknesses in the data can mislead even when each individual statement is accurate.

• A. Publish all raw data so readers can verify the findings themselves
• B. Weigh the public interest of the story against the potential harm to individuals, and minimize personal identifiers where possible
• C. Always include full names to add credibility to the analysis
• D. Share the dataset with advertisers to fund further investigation

• A. Remove the outliers without mentioning them to make the story cleaner
• B. Keep the outliers and disclose how they affect the results, explaining any decisions made about them
• C. Exaggerate the outliers to make the story more dramatic
• D. Delete the entire dataset and avoid publishing the story

• A. Keeping the analysis secret so competitors cannot copy the story
• B. Explaining the data sources, how the data was cleaned and analyzed, and the limitations of the findings
• C. Publishing only the final chart without any explanation
• D. Letting the data speak for itself without any context

• A. Labeling axes clearly and starting a bar chart at zero
• B. Truncating the y-axis of a bar chart so small differences look dramatic, without indicating this
• C. Citing the source of the data beneath a chart
• D. Using consistent colors across related charts

• A. Publish immediately since the agency is an official source
• B. Verify the data independently, check for errors, and contact the agency for comment before publishing
• C. Correct the error privately and publish the flattering version
• D. Ignore the dataset entirely because government data is always unreliable

• A. Publishing the raw dataset without any context so readers can draw their own conclusions
• B. Leading with a clear headline and key takeaway that tells readers what the data shows ✓
• C. Including every chart you created during the analysis to demonstrate thoroughness
• D. Using technical statistical jargon to establish credibility with readers

Why: Readers need to quickly understand the main point of a data story. A clear headline and key takeaway orient the audience before they dive into charts and details, while raw data without context often confuses rather than informs.

• A. It is only required for academic publications, not journalism
• B. It allows readers to assess credibility and understand the limitations of the findings ✓
• C. It makes the story longer, which improves search engine ranking
• D. It prevents other journalists from reusing the data

Why: Transparency about sources and methods lets readers judge how reliable the findings are and what they can and cannot conclude. This builds trust and is a core ethical practice in data journalism.

• A. A pie chart
• B. A line chart ✓
• C. A scatter plot with no time axis
• D. A word cloud

Why: Line charts place time on the horizontal axis and show trends clearly, making them the standard choice for time series. Pie charts show parts of a whole at one moment, and word clouds do not represent quantitative change.

• A. It always makes the data more accurate
• B. It can exaggerate small differences and mislead readers about the size of changes ✓
• C. It hides outliers from the audience
• D. It converts the chart automatically into a percentage chart

Why: When a bar or line chart's axis starts at a value other than zero, small differences can look dramatic. Journalists should label axes clearly and choose scales that represent the data honestly.

• A. Hide the limitations so the story is more compelling
• B. Remove any data points that are uncertain
• C. State the limitations clearly and explain how they affect the conclusions ✓
• D. Present uncertain figures as exact numbers to avoid confusing readers

Why: Honest reporting acknowledges uncertainty and explains its impact on the findings. This helps readers interpret results correctly and maintains the journalist's credibility.