Month: February 2017

1. 1. Tell the “why” and “how”

   Use a descriptive chart title and annotation that not only describe what is being measured rather also why the reader should care and how to read the chart. This will avoid misinterpretations and save time for the chart viewers.

   Example:

   Original title: MSIS degree

   Improved title with note: MSIS degree placement rate. Note: 86% of the MSIS graduates had job placements, which is the highest placement rate when compared to other programs.

   2. Highlight what’s important, tell one story

   Although it is possible to tell 100 of story using a single line chart, it makes much sense to keep the focus on just one story.

   Example: Consider this image, There are 5 products in the chart, and it is not clear what product is the story focusing on. Therefore we must highlight the line that we are focusing on to tell that particular story and keep the rest in context in the background.

   3. Do not use 3D charts

   Studies show that 3D effect reduce comprehension. The extra dimension can hide the visibility of the data, and therefore unable to understand the pattern in data.

   Example: consider this 3D effect image, as we see from the chart, most of the data are hidden, and hence are not easily understandable.

   reference: http://www.eea.europa.eu/data-and-maps/daviz/learn-more/chart-dos-and-donts#toc-0

    

Interactive Visualization enables the display and intuitive understanding of multidimensional data provides a variety of visualization chart types and enables users to accomplish traditional data exploration tasks by making charts interactive. Interactive Visualization implies the use of heat maps, geographic maps, link charts, and a broad spectrum of special purpose visualizations that surround processes that are inextricably linked to an underlying analytics.

Any enterprise at some points starts trending contextual data for making decisions affecting their operations on daily basis. With this trending, the business question which needs to be answered becomes the focus.A visual representation can be done in many ways but here the context matters for better understanding and better decision-making which is provided by Interactive Visualization. Interactive Visualization begins with a data presentation architecture that seeks to identify, locate, manipulate, format, and present data in such a way as to communicate its meaning optimally.

• 5 Key Properties of Interactive Visualization –

1. The Novice User – Even the naive users should be able to examine the data and find all the patterns, correlations and navigate through the visualization easily.
2. Driving Processes – The processes must be well defined. Phase completion should be visually shown and should be real-time.
3. Data Must Tell a Story – Data must tell a story that instantly relates the performance of a business and its assets. The users should be able to select data and change perspective for a better result.
4. Data Correlation – The trends which can be dynamically formed between multiple datasets should be easily found.
5. Prescriptions – The users must be provided with at least some prescriptive analysis. They must also be prompted with the steps to follow to get the desired result.

 

Ref : http://www.forbes.com/sites/benkerschberg/2014/04/30/five-key-properties-of-interactive-data-visualization/#2b128fd744eb

Key Performance Indicators, better known as KPIs are measurements made at certain intervals(weekly, monthly, quarterly, yearly and so on) which provides the business owners an indication of the relative health of the business. Even though it is imperative to consistently measure KPIs in any flourishing business, most small scale business owners ignore it citing the practical difficulty in measuring the same.

Parallels can be drawn between KPIs in business to KPIs in the health sector. It is not just a single instance(in most cases) that triggers concern and treatment from a doctor, but a series of suspicious/abnormal results. Similarly, KPIs in business takes more meaning when the measurements are taken repeatedly and over a period of time.

In business, KPIs are of two categories – Leading and lagging. As the name suggests, leading indicators provide the owners with a glance into the future while lagging indicators are all about the results of previous actions/policies. Leading indicators are useful in predicting the general direction of the business and as such, can be used to alter/continue the course of action depending on the predicted outcome. On the other hand, lagging indicators provide an assessment of the direction in which the business moved in the given period of time.

While useful in its own right, the true potential of KPIs is unlocked when the best of both worlds are combined. For example, a businessman can keep a target revenue for the end of the month and work on achieving it. Leading KPIs can be utilized to assess the progress and depending on the situation, changes can be made if required to achieve the target.

While measuring KPIs, one needs to be prudent in selecting the necessary indicators. Great care must be exerted in the selection of the variables since an increase in number or a seemingly important, but inutile variable could potentially confound the measurements. Research by Drs. Kaplan and Norton came up with a solution for this in the form of their Balance Scorecard, which emphasizes focus on four key areas – Financial, Customer, Internal Business Process, and Innovation and Learning. This framework is meant to align the goals of a business with the strategy and long-term vision.

Source : http://www.business2community.com/small-business/what-are-kpis-and-how-do-you-use-them-01641939#8VF8KMOItt9HgRGI.97

Earlier this week, Hans Rosling a pioneer and one of the leading members in the visualization domain passed away. Hans gave a TED talk in February 2006 and while Hans goes on to talk about how there is a need for the public and private statistical data to be made available to people that need it, the most important take away that I got from his presentation was that society as a whole is more interested in looking at the data from the top most level. We see the world as us against everyone else and countries as first world and third world. People in society never really look and try to understand the data as what it really is about, but instead sees it as what it is shown to us. For example, Hans does a comparison between GBP of countries in the world vs the Child Survival rate. The sub-Saharan region of Africa has the lowest GDP vs Child Survival rates in his data set. If you were to look into the data then you would notice that while the average value is the lowest, the countries that are part of the sub-Saharan region is actually evenly spaced out. Mauritius actually has better statistical value than the average of Latin America. No one would know to look at Mauritius to see why their GBP is so high in the area, but instead they would look and see that the sub-Saharan region has one of the lowest GBP vs Child Survival rates in the world.

Reference: https://www.ted.com/talks/hans_rosling_shows_the_best_stats_you_ve_ever_seen

United Nations maintains an interactive dashboard containing visualizations about world population and related parameters.

The things that I liked about the dashboard are:

• The back/return button at the top left corner of the map is very intuitive since they follow common application norms, such as undo/return on Microsoft applications.
• The icons used for fertility are super likable!

The things I did not like about this dashboard are as follows:

• The color mark used in the map tells which countries have higher and lower population without giving a numeric range to it. Worse yet, even when you hover on a country, there is no tooltip to mention the current population. Ideally, when you think about world population, we would want to know the growth rate for each country. This is probably the second most important data point (first being the current population) when you talk about the population domain. Both these data points are available in the additional parameter section to the user if he/she clicks on a country/region.
• There are four text filters at the bottom of the map, which partitions the world based on development index. When the user clicks on any one of these, the additional parameters get populated for the filter region selected. I would have liked the countries which fall under each of these indexes to be highlighted in the map when each of the filters were clicked. This would have helped the user to understand which countries are falling under them.
• When you click on a country, the map zooms in and its data points are presented in the additional parameter section. I don’t see the zoom feature fulfilling any purpose.
• The tooltips for “Maternal and newborn health” visualization is incorrect and there is no tooltip for “Sexual and reproductive health”.

References:

http://www.unfpa.org/world-population-dashboard

I am rewriting last week’s blog entry with an example to emphasise the importance of the kind of data that can be present ( categorial, ordered, ratio) and how can it be visualized. I found the visualization which shows the Titanic Survivors. The interesting thing about this graph is that it shows the number of individuals as per different categories (dimensions) like Status (survived/perished), Sex (Male, Female), Age (Child, Adult), Class of travel (First, Second, Third, Crew).

Even though the visualization at the first glance looks chaotic I really liked the way in which the author has arranged the dimensions where by they are connected or grouped based on other dimensions. In this way we can get the exact number/percentage for example –  third class female child perished is 1%.  Similarly, if we hover on the category that gives the total numbers for example total crew 40% (885).

Another noteworthy design is the way the “mark” of the chart is separated as it flows from top  to the bottom category.
I want to conclude that the author has done an excellent work in visualizing the dataset which contains different categories.

 

Last Week’s Entry

For this week’s blog entry I would like to summarize the “what data”. As I am starting to work on my projects I wanted to look at the data I have collected so far.   

There are two types in which the data is stored

1. Table Data – table data has attributes and rows. Each column/field/attribute explains what type of data is present in the row.  
2. Metrics Data – metric data is the  has 2 or more dimensions which represent a data point.  This more useful information for analytical purpose.

Three broad kinds of data –

1. Categorial data – the data is represented in categories, categories of the movie like humour, horror etc. We cannot do calculations solely on this data. 
2. Ordered Data – the data is presented in terms of ranks, we can definitely say which city is better than another but we cannot  elaborate on how much it is better than the other by this kind of data.
3. Ratio Data – This data has numbers and we can do calculations as the data is quantifiable. For example We can clearly say milage of car A is better that of car B by 6 miles/gallon.

By deciding which of the categories does my data fall under I can decide on why part of the analysis.

Last week, we all completed the exercise 2 and for some of us this was the first dashboards we created. Even though using tableau to create a chart or a graph is super easy, analyzing it to get to the results you need is a time-consuming task and requires lot of iterations. So, after I completed my first dashboard I tried to analyze if that was the best I could do. This got me to research on the methods and approaches to designing great dashboards. Upon that I came across this article about “Designing and Building Great Dashboards”

“Different people in the company ask for different data to be displayed and soon the dashboard becomes hard to read and full of meaningless non-related information.” (SMITH, 2015) So, focusing on these high-level design rules help us to create a dashboard that is worth the time and effort we put in to designing it.

Rule 1: WHO ARE YOU TRYING TO IMPRESS?

The most effective dashboards target a single type of user and just display data specific to that ‘use case’.

 

Rule 2: SELECT THE RIGHT TYPE OF DASHBOARD

Dashboards are of different types and each of them is used for performing a specific purpose.
The types of dashboards are Operational, Strategic / Executive and Analytical dashboards.

 

Rule 3: GROUP DATA LOGICALLY – USE SPACE WISELY

Grouping data is very important to get the dashboard right. Either grouped by department or functional area.

 

Rule 4: MAKE THE DATA RELEVANT TO THE AUDIENCE

Ensure that the data you display on the dashboard is relevant to the users. The components should always be designed thinking about the scope and for data to reach of your users.

 

Rule 5: DON’T CLUTTER YOUR DASHBOARD – PRESENT THE MOST IMPORTANT METRICS ONLY

Whether it is useful and useless information added to fill the dashboard cluttered dashboards don’t give the impact. It often takes away the focus from the important messages.

 

Rule 6: HOW OFTEN DOES THE DATA REALLY NEED TO BE REFRESHED?

For dashboards that are interactive, we always have to keep in mind that the data keeps changing and so the dashboard has to be updated.

https://www.geckoboard.com/blog/building-great-dashboards-6-golden-rules-to-successful-dashboard-design/#.WJ9m1rYrKRs

Heat maps use color variance for data visualization. They are intensive used for displaying variance between different variables, displaying any particular pattern between them and if any correlation are present between the variables.

• The rows and columns of a table form the matrix structure of the heatmap. Each cell of the matrix contains color coded data or numerical data which is displayed on a color scale. The matrix data represents the relationship between the variables of the row and column associated.
• A legend should be given alongside the heatmap for better understanding of it. Numerical data requires a color scale which has different colors blending into one another to show variance of high and low in the associated data. While categorical data is color coded.
• Heat map uses the color differences to display changes in value, hence it should be used to give a more generalized view of the numerical data. Heatmap should not be used to display sensitive data which needs to be represented accurately.
• Heat maps are best used to show changes in values over time. Any column of row can be used to denote the time changes.
• The colors in the heatmap should be chosen carefully as the difference must be visible immediately to the human eye. Rainbow color schemes are highly used as humans can perceive more shades of those colors. Grey color scales must be avoided as they are difficult for perception.
• The best use of heat maps are done to show temperature changes in a city or town over months or years or to depict the hottest and coolest places to stay.

Source: http://www.datavizcatalogue.com/methods/heatmap.html

https://www.bloomberg.com/news/articles/2015-12-03/electric-cars-can-t-take-the-cold

This week we had an interesting discussion in class, about when to use heat maps and how to interpret them.

I came across this heat map. The claim is, Since electric cars generate power less efficiently as the temperatures drop, they are sold more on the West coast than other regions of USA. And obviosuly, there are other reasons like West Coast being more technology savvy than the rest, also adds up to the sales.

About the visualization –

• The heat map  is U.S electric vehicle sales by region
• On observing we see, that there are 4 patterned boxes for 4 regions and California is in light blue
• On the first glance, one state and other regions seems confusing
• It takes time to interpret the heat map

The Underlying meaning

• The visualization depicts the 4 regions and the sales made in each region in September
• What they have tried to show is, even on the West Coast, California sells the highest number of electric cars.
• The number of cars sold in California, is greater than those combined for Midwest, Notheasrt, South and the remaining West region.
• But, this could be shown through a bar graph as well, comparing sales in California with the rest. That would have been easier to interpret in the first look.

Apart from temperature, factors like population should also be considered. California is the most populated state , as per the region available, so that could be one contributing factor as well.

 

References –

https://www.bloomberg.com/news/articles/2015-12-03/electric-cars-can-t-take-the-cold

After working on the visualization exercise 2 I really got to appreciate the amount of effort taken by designers to convey the meaning of the charts effectively. Tools such as Tableau give us chart suggestions to choose from depending on the measures and dimensions we choose. The advantage of this is translating the chosen attributes into a visualization is convenient for anyone even without data skills. But, this doesn’t necessarily serve the purpose of the insights that you would want your visualization to communicate. (Going back to our viz exercise 2, it may not be enough if you compare MSIS with other degrees it would be more effective to show MSIS is a better program to choose because it offers a more stable mid-career pay with less uncertainty.)

Steering away from viz exercise 2, on a general note when you’re trying to think through the purpose of your visualization you could start of by answering two questions:

1. Is your data trying to show ideas or statistics? – this leans more towards the underlying data rather than the form of visualization. E.g. for idea – organization structure chart. E.g. for data-driven – Revenue growth for last five years.
2. Are you declaring something or exploring something? – An example of declaring something is if you want to project the quarterly sales of 2016 to your manager with the available sales data. However, if you want to understand why the sales performance is lagging. You suspect that there is a seasonal drop and want to prove the same with a quick visual which become an exploratory kind of visualization where form takes priority over the available information.

Now that you know what you want to communicate to your audience the next step would be to analyze the type of the chart to choose (i.e. when to use a bar chart over line chart). Also, thinking through an effective way of using the color palette. Other aspects could be the ordering in a bar graph, starting your x axis from zero, etc. While following, these chart making rules you should always make sure your chart is communicating your claim/insight clearly.

Reference – https://hbr.org/2016/06/visualizations-that-really-work