{"id":12327,"date":"2020-11-27T16:13:40","date_gmt":"2020-11-27T15:13:40","guid":{"rendered":"https:\/\/www.codemotion.com\/magazine\/?p=12327"},"modified":"2022-07-08T15:38:58","modified_gmt":"2022-07-08T13:38:58","slug":"data-visualization-engagement-rules","status":"publish","type":"post","link":"https:\/\/www.codemotion.com\/magazine\/ai-ml\/big-data\/data-visualization-engagement-rules\/","title":{"rendered":"5 Rules of Engagement When it Comes to Data Visualization"},"content":{"rendered":"\t\t\t\t
\n\t\t\t\t
\n\t\t\t\t\t\t
\n\t\t\t\t\t\t\tTable Of Contents\t\t\t\t\t\t<\/div>\n\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t
\n\t\t\t\t\t\t
  1. Number 1: What\u2019s the point?<\/a>
  2. Number 2: Know your options<\/a>
  3. Number 3: Understanding your data<\/a>
  4. Number 4: Aesthetic of data visualization<\/a>
  5. Number 5: Introduce your work to criticism<\/a><\/ol>\t\t\t\t\t<\/div>\n\t\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\r\n\r\n\r\n

    So you\u2019ve been out, you\u2019ve conducted the surveys<\/a>, you\u2019ve collected a giant pot of data<\/span> that unequivocally demonstrates your position and now it\u2019s time to present. How do you do it? Well you could just dump the raw spreadsheet<\/span> at the desk of your manager or whoever it may be, declaring \u2018it\u2019s all there!\u2019, triumphantly. Or maybe organize a teleconference<\/a> or web meeting<\/a> where you rattle off data<\/span> points hour after hour.\u00a0<\/p>\r\n\r\n\r\n\r\n

    Of course, these are intentionally ludicrous suggestions. The key to communicating the meaning of data<\/span> to those who need to understand<\/span> it is to provide an effective, compelling, and easy-to-interpret visualization<\/span>. In the data-driven culture<\/a> in which we live, being able to communicate data<\/span> interpretations<\/strong> in this way is an essential skill for any business person to have.<\/p>\r\n\r\n\r\n\r\n

    Let\u2019s outline five basic rules of engagement<\/span> for data visualization<\/strong><\/span>.<\/p>\r\n\r\n\r\n\r\n

    Number 1: What\u2019s the point?<\/h2>\r\n\r\n\r\n\r\n

    Don\u2019t worry you\u2019re not seeing signs of an existential episode, our first point indeed is – what is the point? You ought to determine what point you are attempting to make with your data<\/span><\/strong> before you take the step to visualize<\/span> it. Without this step things will get messy and confusing.<\/p>\r\n\r\n\r\n\r\n

    \r\n
    \"Data<\/figure>\r\n<\/div>\r\n\r\n\r\n\r\n

    Consider the above infographic. What is the point? What message<\/span> is the analyst who built this graph<\/span> trying to tell us? If there ever actually was one, it has been lost to the sea of potential interpretations an audience<\/span> member could impose onto it. It is necessary to minimize this risk of alternative interpretations of your data<\/span> visual if you want to deliver a concise message<\/span>.<\/p>\r\n\r\n\r\n\r\n

    Ask questions, get answers<\/h3>\r\n\r\n\r\n\r\n

    A way to do this is to identify particular questions<\/strong> that your data<\/span> can provide an answer to. For example, imagine you have collected a large sample of data<\/span> regarding the varying success of a particular ad campaign across numerous media<\/span>. You now have the information necessary to assess the relative success of the various media<\/span>. Now ask your questions.<\/p>\r\n\r\n\r\n\r\n

    You can ask this data<\/span>, which is the most successful advertising medium<\/span> of all? Is it TV, social media<\/a>, enterprise e-commerce<\/a> platforms, radio, or something else?\u00a0<\/p>\r\n\r\n\r\n\r\n

    Think of this as a surface-level, fundamental question where there is a clear and direct association between the data<\/span> and the interpretation<\/span> applied to it. For example, perhaps the initial data<\/span> shows that social media<\/span> is the most effective of the channels due to its superior success rate against the rest. There is no further analysis<\/span> to be done in order to answer the question we set out to answer. No reasonable person could deny the interpretation<\/span> that social media<\/span> is superior.<\/p>\r\n\r\n\r\n\r\n

    Thereafter, you may then explore secondary interpretations<\/strong> that can be inferred<\/span> from the data<\/span>. Maybe you also collected personal data<\/span> from successfully onboarded customers<\/span> within each of the media<\/span>. Then, you can assess the success rate of particular media<\/span> among specific demographics<\/span>. This would constitute an additional layer<\/span> of analysis<\/span> and therefore your visualization<\/span> will need to be more sophisticated.<\/p>\r\n\r\n\r\n\r\n

    Layers of data<\/span> interpretations<\/strong> can build up in this way, so – depending on what question you\u2019d like to see answered and demonstrated to your audience<\/span> – you will need to find the most appropriate way of presenting these layers<\/span> of information.<\/p>\r\n\r\n\r\n\r\n

    Generally, deeper levels of analysis<\/span> will require a more sophisticated visualization<\/span> than more direct data<\/span> sets.<\/p>\r\n\r\n\r\n\r\n

    Number 2: Know your options<\/h2>\r\n\r\n\r\n\r\n

    When it comes to data<\/span> presentation<\/span><\/strong>, you have a variety of options. Try and get familiar with the various ways data<\/span> can be visualized. Let\u2019s just very quickly take a look at a few of these ways, and how they are best used.<\/p>\r\n\r\n\r\n\r\n

    Data visualization with bar graphs<\/h3>\r\n\r\n\r\n\r\n

    Perhaps the most familiar of data<\/span> visuals is the bar chart<\/span><\/strong> or bar graph<\/span>. It is used when measuring the variability of a single factor between different categories.<\/p>\r\n\r\n\r\n\r\n

    \r\n
    \"Data<\/figure>\r\n<\/div>\r\n\r\n\r\n\r\n

    The above provides a rudimentary example of a bar chart<\/span>. As described above, you can see that there is one variable<\/span> that is being measured (number of people) and the graph<\/span> provides a mode of comparison between a given set of categories (favorite type of movie) listed across the bottom.<\/p>\r\n\r\n\r\n\r\n

    The bar graph<\/span> is useful when making a basic comparison between categories in order to determine which is the best or worst option.\u00a0<\/p>\r\n\r\n\r\n\r\n

    Data visualization with histograms<\/h3>\r\n\r\n\r\n\r\n
    \r\n
    \"Data<\/figure>\r\n<\/div>\r\n\r\n\r\n\r\n

    The above image is an example of a histogram<\/span><\/strong>. It is similar to a bar graph<\/span> in aesthetic, but the type of data<\/span> used with histograms<\/span> differs. Whereas in a bar graph<\/span> the categories are distinct, in a histogram<\/span> the categories are ranges from a continuous data set<\/span>. It consequently demonstrates proportional prolificacy of the separate ranges.<\/p>\r\n\r\n\r\n\r\n

    The continuous nature of the data<\/span> is indicated by the lack of gaps between bars.<\/p>\r\n\r\n\r\n\r\n

    You should use a histogram<\/span> if you are looking to compare two continuous sets of data<\/span> and want to categorize groups<\/span> within one of those sets.<\/p>\r\n\r\n\r\n\r\n

    Data visualization with line graphs<\/h3>\r\n\r\n\r\n\r\n
    \r\n
    \"Data\r\n
    Data Visualization: Line Graphs<\/figcaption>\r\n<\/figure>\r\n<\/div>\r\n\r\n\r\n\r\n

    A line graph<\/strong> shows the differences in one variable (y-axis<\/span>) in relation to an independent<\/span> continuous variable<\/span>; usually time (x-axis<\/span>). They are useful for identifying trends and patterns in data<\/span> and can be used to show how one independent<\/span> variable affects multiple things at a time.\u00a0<\/p>\r\n\r\n\r\n\r\n

    For example, you could use the same graph to show the amount of annual rainfall in different areas of the country, with each area being represented by a separate line. Line graphs should be considered if you have a lot of data<\/span> to present.<\/p>\r\n\r\n\r\n\r\n

    Data visualization with pie charts<\/h3>\r\n\r\n\r\n\r\n

    Pie charts<\/span><\/strong> are a way of demonstrating the total distribution of all your data points<\/span><\/strong> among different categories. They should be used as an alternative to a bar graph<\/span> if you are more interested in showing the proportionality<\/strong> of your results than the relative amounts. For example, it may be useful to employ a pie chart<\/span> if you are looking to personalize marketing<\/a> techniques and need to assess the proportion of your customers<\/span> who are positively affected by personalized ads.<\/p>\r\n\r\n\r\n\r\n

    Summary: which chart for your data visualization<\/h3>\r\n\r\n\r\n\r\n

    These four are perhaps the most common, but there are many more and it may be worth your while to research<\/span> into the different options. Your decision<\/span> as to what is the best visualization<\/span> format to use will be informed by the characteristics of your data<\/span>.\u00a0<\/p>\r\n\r\n\r\n\r\n

    Number 3: Understanding your data<\/h2>\r\n\r\n\r\n\r\n

    Right, so you know what questions you want your visuals to answer, and we\u2019ve looked at a few of the available options for representing those answers. Now you have to make a decision<\/span>. This will come down to the characteristics of the data<\/span><\/strong> and the variables<\/span> you need to display<\/strong>.<\/p>\r\n\r\n\r\n\r\n

    Types of variables<\/h3>\r\n\r\n\r\n\r\n

    Variables<\/span> come in two major types – independent and dependent variables<\/span><\/strong>. Broadly speaking, the independent variables<\/span> are the things that we change and the dependent variables<\/span> are the things that we measure. We are measuring the effect of the independent variables<\/span> on the dependents.<\/p>\r\n\r\n\r\n\r\n

    \r\n
    \"Data<\/figure>\r\n<\/div>\r\n\r\n\r\n\r\n

    For example, if we are assessing marketing communication tools<\/a>, the tools<\/span> themselves would be independent variables<\/span>. Their success rates would be the thing we measure and, therefore, the dependent variable.<\/p>\r\n\r\n\r\n\r\n

    It is important to make these distinctions as different visualization methods will suit certain arrangements better. Bar graphs and pie charts, for example, are great ways to show proportions between independent categories against a dependent variable.<\/p>\r\n\r\n\r\n\r\n

    Types of data<\/h3>\r\n\r\n\r\n\r\n

    Data can be either qualitative or quantitative<\/strong>. But within each of these there are multiple different forms data can take.<\/p>\r\n\r\n\r\n\r\n

    Qualitative<\/h4>\r\n\r\n\r\n\r\n

    Qualitative data is data that is subjectively judged or grouped into categories. Specific types include:<\/p>\r\n\r\n\r\n\r\n