"A time series is a sequence of values, usually taken in equally spaced intervals. […] Essentially, anything with a time dimension, measured in regular intervals, can be used for time series analysis."
"Calculating the percent change between two percentages is not completely inaccurate, but it can be very misleading. Instead, you should use the absolute change when you are working with percentages and want to show the difference between two points in time." (Andy Kriebel & Eva Murray, "#MakeoverMonday: Improving How We Visualize and Analyze Data, One Chart at a Time", 2018)
"Data analysis is more than crunching numbers; it is about finding
insights, identifying the unknown unknowns, and presenting the data in a simple
yet deep enough way so that your audience can understand your insights and make
decisions."
"Heat maps are effective visualizations for seeing concentrations as well as patterns. Adding time series to a heat map can also reveal seasonality that may not be obvious otherwise."
"Ideally, the charts are designed in a way that gives your audience clarity and lets them understand the key insights very quickly. Color choices, highlighting, annotations, and other ways of drawing attention to your findings help in the process. By leaving white or blank space around your charts, you are able to keep the focus of your audience on the key message rather than distracting or confusing them." (Andy Kriebel & Eva Murray, "#MakeoverMonday: Improving How We Visualize and Analyze Data, One Chart at a Time", 2018)
"Plotting numbers on a chart does not make you a data analyst.
Knowing and understanding your data before you communicate it to your audience
does."
"Ranks do not explain how much one item varies from another. Ranked
data is ordinal; that is, the data is categorical and has a sequence (e.g., who
finished the race first, second, and third). That’s it! Ranked data can be used
for showing the order of the data points. […] When working with ranked data,
you cannot make inferences about the variance in the data; all you can say with
certainty is which item is ranked higher than the others, not how much higher."
"Simplicity for data visualization often focuses on minimizing the number of elements that do not add value to your display. These include borders, gridlines, axes lines, and boxes, which can easily distract from your core message. This recommendation also relates to the information itself. You should strive to create a visualization that focuses on specific aspects of the data, rather than including all fields and metrics but not saying much about any of them."
"Simplicity in design can be recognized in visualizations that are clear, easy to understand, uncluttered, and impactful. Nonessential items are removed from these visualizations so that the data stands out, giving it space and removing distractions. Simplicity in design pays careful attention to the overall layout and positioning of individual components, the balance of charts and text elements, and the choice of colors, fonts, and icons, as well as the clarity with which all of these elements communicate to the audience."
"Taking an average of an average (the original percentage) does not result in a weighted average, which takes into account the sample size […]."
"[…] the drawback of the box plot is that it tends to hide the values due to its design."
"To become a great data analyst, you must be able to identify and deal with incomplete data and work to identify the data quality and accuracy issues in a data set."
"Using a question as a title is a great way to guide the audience. The question helps you ensure that your charts respond directly to the question and when they do not, you can remove them. And that is the main point: You need to answer the question. If the data is not conclusive, say so. Give an explanation that relates back to your title and close the loop so that your audience is informed and gets the complete picture included in your analysis."
"Visually plotting time series data against a point in time reveals patterns relative to that period, thus allowing the reader to understand growth and decline before and after the given point in time."
"When using indexes in a data set, using an average aggregation is appropriate as long as you only use it at the individual region, month, and visitor type level (i.e., the lowest granularity of the data). You cannot use an average of the average to represent the total." (Andy Kriebel & Eva Murray, "#MakeoverMonday: Improving How We Visualize and Analyze Data, One Chart at a Time", 2018)
"When you are exploring your data, look for alternate views
of the data; you just may find a more interesting insight."
No comments:
Post a Comment