Graphical Perception Essay

Graphical representation of data plays a key role in understanding the statistical data. When provided with bulk data it is easier to interpret and make judgement by looking at graphs rather than huge spreadsheets. There are various graphs available to project data , the biggest challenge is to decide which graph will be the most effective for the given data set. Important things to keep in mind are , we should not clutter the data or miss any important details. The Elements of graphing data shows Cleveland’s method to present data in a meaningful way also helps in data analysis and data exploration.

Cleveland and McGill placed the foundation of how data visualization is best done. They did a lot of experiments to understand and relate human interpretation to understand graphs. Graphical perception is nothing but analyzing the information encoded by graphs.

knitr::include_graphics("/Users/parmars//Documents/Kavita/Semester 4/Anly 512/image1.jpeg")

Figure 1 shows the usage of Weber’s law of lengths. The bars A and B are of different lengths. But the difference is so small that it is difficult to visualize. But at the same time if we look at the way the bars ace visualized in the second section we can clearly see the white gap for the Bar A is larger than the white gap of Bar B(considering both the bars are placed with equal lengths). Since the white is larger in Bar A it clearly shows that A is smaller as compared to B. This classification clearly explains how important it is to choose the right graph for data interpretation.

knitr::include_graphics("/Users/parmars//Documents/Kavita/Semester 4/Anly 512/image2.jpeg")

Figure 2 shows 8 curves. Our task is to figure out which one dips the lowest. Are all the 8 curves dipping at the same minimum point. Figure 3 shows the exact same curves but with gridlines. This makes us understand the graph better as we can mark the dip points via grid lines. For comparison sake we could have also plotted all the curves in the same graph , but that would clutter the graph and it would have been very difficult to understand.

We consider graphs as a method for data exploration. But it is an ongoing process and we may have to replot based on new findings This is a multi iterate process. A good final graph should be a good representation of the data which will answer all our questions related to the data.

References :-

1-W. S. Cleveland and R. McGill. Graphical perception: Theory,experimentation, and application to the development of graphical methods. J. Am. Statistical Assoc., 79:531-554, 1984. 2-Jeffrey Heer, Michael Bostock .Crowdsourcing Graphical Perception: Using Mechanical Turk to Assess Visualization Design ACM Human Factors in Computing Systems (CHI), 203-212, 2010 3-Good Graphs: Graphical Perception and Data Visualization: The Win-Vector LLC data science blog 4-Zubiaga, Arkaitz and MacNamee, Brian. (2016) Graphical perception of value distributions : an evaluation of non-expert viewers’ data literacy. Journal of Community Informatics, 12 (3). ISSN 1712-4441 5-J. Talbot, V. Setlur, and A. Anand. Four Experiments on the Perception of Bar Charts. In IEEE Transactions on Visualization and Computer Graphics, Volume: 20, Issue: 12, Dec. 31 2014.