Data Visualization: MCQ Component 1 PDF

MCQ Component 1 Basics Theoretical concepts - Need for Data visualisation - Steps in DV - Points to keep in mind - Principles of design - Types of charts and use   Humans process visual information easily  Decision makers do not understand complex anal...

MCQ Component 1 Basics Theoretical concepts - Need for Data visualisation - Steps in DV - Points to keep in mind - Principles of design - Types of charts and use   Humans process visual information easily  Decision makers do not understand complex analytics  Large information in small area Why DV is needed?  Interactivity can be introduced via dashboards  Relationships and comparisons, Trends and Focus areas  Improved insights  Data visualisation & Visual Analytics Infographic s Data visualisatio n Calculate – Mean of x,y – Variance of x,y – Correlation between x,y – Linear regression – Plot x vs y for each data set Use SPSS, Excel Descriptive or Graphical? Use both DA and VA go hand in hand https://www.visual-analytics.eu/faq/ Steps for DV  You need to prepare a dashboard for the CMO of a company. The company maintains the historical records of sales. How will you go about it? 1. Acquire Example : 2. Parse Steps for 3. Filter DV 4. Mine 5. Represent 6. Interact  1 Context, Audience , Goal 2 5 Keep in Feedbac k Audience Attention mind 4 3 Tell a Visual story type   Context :  What business question needs to be answered? 1  Background information relevant to what needs to be communicated  Audience :  Context,  Who is the audience? Audience biases Audience,  What information is relevant to audience?  Goal How will the audience use the DV  Goal :  Visualisation to business question  Information needed  What question is more important?  2 Audience attention Z-Pattern of layout  2 Audience attention  Follow 3 important concepts 2 1. Pre-attentive attributes Audience 2. Tufte Guidelines attention 3. Gestalt Principles  1. Pre-attentive attributes 2  Colour  Form  Movement  Spatial Positioning Audience attention  2. Tufte’s Principle [A] Graphical Integrity : [B] Maximise data Ink Visual representation of data 2 must tell the TRUTH Lie Factor which can be calculated by dividing the size of the effect shown in the graphic by the size of the effect in the Audience data. attention P1 :Numbers should be proportional to numerical quantity P2 : Clear Labels and Label important events P3 : Show data variation, not design variation P4 : Use standardised units instead of nominal units P5 : No of dimensions should not exceed No of variables carrying info 2. Tufte’s Principle [C] Reduce chart junk [D] High Data density : Shrink Principle [E] Small Multiples  A- P3 A- P5 What is the Lie factor ? 2001 : 3.58 cm 2002 : 2.92 cm 3. Gestalt Principles  Use Size shape colour to force the viewer to group 1.Similari objects. This can be a great ty tool to show users clusters Since the eye naturally 2.Continuati follows a line, placing items on in a series in a line will Examples naturally draw the eye from one item to the next. Use it to lead the viewer, when you show a partial image fading off the user’s screen in order to show them 3.Closure that there is more to be found if they swipe left or right. Without a partial image, i.e., if only full images are shown, the brain doesn’t immediately interpret that there might be more to be seen, and therefore your user is less likely to scroll (since closure is already apparent).  4.Proximit y By putting like things closer together, with space in between each group, the viewer will immediately pick up on the organization and structure you want them to perceive. Put two things you want users to relate together near. If you want them to not relate them, put them far.  5.Ground/Figure Use when want to highlight a focal point, particularly when it is active or in use—for example, when a modal window pops up and the rest of the site fades into the background, or when a search bar is clicked on and the contrast is increased between it and the rest of the site.  6.Symmetry Use symmetry to bring about stability and order.  7. Common fate Objects moving in same direction cause the viewer to believe there exists a relation between them and they all have one common fate or goal.  8. Simplicity – good figure Objects moving in same direction cause the viewer to believe there exists a relation between them and they all have one common fate or goal.  9. Past experience People often try to use past experiences to perceive what they see. Use this to you advantage. Example : people will relate the above to traffic signals and also relate green to go and red to danger.  3 Visua l Type   Storytelling has 5 parts 5 1. Big idea : Conveyed when you see the visual 2. Beginning : Background 3. Middle : Important events Storytelling 4. End : Actions prescribed & 5. Narrative framework : based on audience Feedback The last part is always take feedback from 1. Colleagues 2. Client 3. Third person And rework  Good or bad ? This dashboard example from Wave is a great example of how a simple structure can be used effectively to display information. The dashboard looks visually appealing but does not draw attention away from the data itself by using distracting and unnecessary graphics. The colour scheme used throughout the dashboard helps it to look bright, clean and clear. The amount of data presented in each graph is sufficient enough to  Another of our good dashboard examples comes from Wufoo  This is a great example of a dashboard which allows the user the ability to drill down and gain a much deeper insight as they make the way down the dashboard.  The dashboard has a flow structure which allows the user to select the area of interest and then allows then to see key metrics in terms of variables such as country, region and city.  The dashboard also effectively incorporates filters which allow the user to view data based on time scales such as day, month, year or last 12 months.  This London City dashboard attempts to present far too much information in a small space and has ended up looking extremely cluttered and distracting.  There are too many distracting colours, many of which serve no real purpose and this draw focus away from the data itself.  There are also far too many different types of data visualization on the dashboard with no clear linkage between them.  The main problem with this dashboard from Dundas is that the same colour scheme is used for both the pie charts at the top and the bar charts at the bottom.  On first glance the user therefore assumes that there is some kind of link between the two and associates the coloured sections of the pie charts with the respective coloured bars below. In fact the two sets of information are not the same and therefore the colour scheme used only complicates things further.  A further fault is the use of 3D effects on the pie charts at the top. In an attempt to make the charts look more visually appealing, the creator has actually made it more difficult for the user to read and interpret.