Scribd Logo
Upload

Welcome to Scribd!

  • 50 views

    The Data Literacy Cheat Sheet: Charts: Which One Should You Use?

    Uploaded by

    deskos
    This document provides guidance on selecting appropriate charts and averages to visualize data, as well as tips for questioning and interpreting data visualizations. It discusses choosing between bar charts, line charts, scatter plots, and other chart types depending on the type of data being displayed. The document also explains when to use the mean, median or mode average and how to identify weighted averages. Finally, it provides best practices for visualizing trends, correlations and avoiding biases when analyzing data.

    Copyright:

    © All Rights Reserved
    Download as PDF, TXT or read online from Scribd

    The Data Literacy Cheat Sheet: Charts: Which One Should You Use?

    Uploaded by

    deskos
    50 views2 pages
    This document provides guidance on selecting appropriate charts and averages to visualize data, as well as tips for questioning and interpreting data visualizations. It discusses choosing between bar charts, line charts, scatter plots, and other chart types depending on the type of data being displayed. The document also explains when to use the mean, median or mode average and how to identify weighted averages. Finally, it provides best practices for visualizing trends, correlations and avoiding biases when analyzing data.

    Original Title

    data-literacy-cheat-sheet

    Copyright

    © © All Rights Reserved

    Available Formats

    PDF, TXT or read online from Scribd

    Did you find this document useful?

    Is this content inappropriate?

    This document provides guidance on selecting appropriate charts and averages to visualize data, as well as tips for questioning and interpreting data visualizations. It discusses choosing between bar charts, line charts, scatter plots, and other chart types depending on the type of data being displayed. The document also explains when to use the mean, median or mode average and how to identify weighted averages. Finally, it provides best practices for visualizing trends, correlations and avoiding biases when analyzing data.

    Copyright:

    © All Rights Reserved
    Download as PDF, TXT or read online from Scribd
    Download as pdf or txt
    50 views2 pages

    The Data Literacy Cheat Sheet: Charts: Which One Should You Use?

    Uploaded by

    deskos
    This document provides guidance on selecting appropriate charts and averages to visualize data, as well as tips for questioning and interpreting data visualizations. It discusses choosing between bar charts, line charts, scatter plots, and other chart types depending on the type of data being displayed. The document also explains when to use the mean, median or mode average and how to identify weighted averages. Finally, it provides best practices for visualizing trends, correlations and avoiding biases when analyzing data.

    Copyright:

    © All Rights Reserved
    Download as PDF, TXT or read online from Scribd
    Download as pdf or txt
    of 2

    The Data Literacy Cheat Sheet

    Charts: Which one should you use?

    Comparing multiple values Displaying the composition of a whole Showing the relationship between
    sets

    Circular
    Column Bar area Area Waterfall Pie
    Line Scatter

    Bullet Stacked
    Line Scatter Stacked
    bar column

    Showing distribution of values Analyzing trends

    Column Bar Line Scatter Column Line Dual-axis-line Bubble

    REMEMBER! Averages:
    The five characteristics Which one should you use?
    of good data:
    An average, also referred to as a “Measure of central tendency”, is a value that attempts to
    identify the central position within a set of data. Mean, Median and Mode are types of
    Credible
    average.
    Actionable
    MEAN Does your data have a continuous distribution that’s relatively symmetrical? Use MEAN
    Unbiased
    (often referred to as just ‘average’).
    Statistically relevant
    MEDIAN Does your data contain significant outliers? Use MEDIAN - it's less influenced by this.
    Easy to interpret
    MODE represents the most common value in a dataset. If you’re dealing with Nominal data
    (non-numeric categories like “industry vertical”), MODE is the only appropriate average to use.

    CHARTING TIPS
    Weighted average
    A weighted average is a type of MEAN, where some values in the data set are given more influence than others. Each value to
    be averaged is given a weight, representing the importance of that value in the average.
    Weighted averages are important when you are dealing with frequencies or distributions, or when working with data that’s
    unequal in some way.

    Common cognitive biases


    CONFIRMATION BIAS SELECTION BIAS
    The tendency to search for results that confirm your preconceptions. The tendency to select data for analysis that is not
    properly random.
    OBSERVATION BIAS
    The tendency to see what we expect (or want) to see in results. SAMPLING BIAS
    (A type of Selection bias) The tendency to collect a
    FUNDING BIAS
    sample of data in such a way that some members of
    The tendency to support the interests of a study's financial sponsor
    the population are less likely to be included than
    or business.
    others.
    How to question data

    SOURCE SCALES FILTERS TIMEFRAME GAPS EXCESS


    Do you know Are the scales of Have any specific What is the date Are there obvious Is there anything
    where the data each axis clear filters been applied range for the omissions to the presented that's
    came from? and effective? to the data set? presented data? data set? not relevant?

    UNIT (S) LABELS ACTIONABLE TREND PATTERNS DIMENSIONS


    Is it clear what Is the data clearly Can the insights Is it trending Are there cyclic Is the data segmented
    the data in the titled and labelled, presented be used upwards, patterns (e.g. into clear, meaningful
    chart represent? in a descriptive in an actionable downwards or seasonality) in the dimensions, e.g.
    way? way? flat? data? “Pricing plan”?

    CHARTING TIPS

    Should I truncate the Y axis?


    Truncating the Y axis (i.e. not starting at zero) is controversial • If you need to emphasise small changes and trends in data,
    in the world of data visualization. It can be misinterpreted and consider showing a chart of the change rather than absolute
    has been used to mislead consumers in a number of cases. In numbers.
    general, it’s not a good idea unless you can clearly show that
    NEVER truncate the Y axis when:
    the axis doesn’t start at zero.
    1. Using a bar or column chart (Bar charts should start at zero).
    • It can be acceptable to truncate when you’re displaying data 2. The information is likely to be mis-interpreted.
    that would never feasibly reach zero anyway, e.g. global 3. It doesn't help in some way with understanding the chart.
    average temperature.

    Data correlation
    Correlation is a statistical relationship between two data sets. Correlation can have a numeric value on a scale from -1 to 1.
    A POSITIVE correlation is present when both values increase together, whereas in a NEGATIVE correlation, one value decreases
    as the other increases.

    e
    ativ osit
    ive
    neg ct p on
    fect n tion e CHARTING TIPS
    Per relatio rela rf
    Pe relati
    Name cor
    cor No cor Correlation is not causation!
    -1 -0.5 0 0.5 1 A strong correlation between two data sets
    Value Negative correlation Positive correlation does not necessarily mean that one thing
    causes the other (causation). There could
    be other reasons the data has a strong
    Typical correlation.
    Chart

    Glossary
    QUALITATIVE DATA is descriptive - it describes something, e.g. Reason for customer cancellation.
    QUANTITATIVE DATA is always numerical (involves numbers), e.g. Revenue lost from customer cancellations.
    DISCRETE DATA can only take certain values (like whole numbers), e.g. Number of customers churned.
    CONTINUOUS DATA can take any value, within a given range, e.g. Customer churn rate.
    CATEGORICAL DATA can be sorted, according to defined groups or categories, e.g. Industry vertical.
    STATISTICAL SIGNIFICANCE is when the observed outcome of an experiment is unlikely to have occurred due to chance.
    This is an important factor when running multi-variant (A/B) tests on your product or website.
    Determining statistical significance can be complex. We recommend using a free tool such as AB Testguide: https://abtestguide.com/calc

    More resources
    The Data Viz Checklist by Stephanie Evergreen: http://stephanieevergreen.com/wp-content/uploads/2016/10/DataVizChecklist_May2016.pdf
    Graph Choice Chart by Tuva Labs: https://tuvalabs.com/static/documents/Graph_Choice_Chart.pdf

    Try the World's Best Subscription Analytics App for FREE at chartmogul.com

    You might also like