Conclusion
A Quick Review...
Data visualizations come in various types.
Drawing with data involves using the raw or processed data to create visualizations.
Drawing with concepts involves using ideas and logic (or illogic) to create visualizations.
Drawing with rules involves using extrapolated rules (from empirical observations) to create visualizations. (These rules do not have to be drawn from "data," but for our purposes, it is. Rules may be drawn purely from the imagination.)
Plus...
There are many other types of data visualizations. There are many that are 3D and 4D (with changes over time). There are many that are highly specialized to particular domain fields.
"Big Data" and Data Visualizations: The current chatter surrounds "big data" to understand patterns and sense-make. The analysis of big data enables the revelation of unanticipated relationships (relationships do not have to be known a priori). They may reveal unanticipated informational blind spots. With "big data," correlations trump causation; in other words, certain indicators may correlate with certain phenomena, and those indicators will be used for understandings and predictions even if the underlying causation is not clear
.
Big data sets themselves do not have to be as precise or clean as in a context of scarce data. This means that there may be cells with missing information in a big data set that is still used for analysis because the thought is that the massive amounts of data will make up for potential gaps. However, big data may lead to mistakes in correlations. Big data is prone to false indicators of correlation, but these apparent correlations may be only artifacts of the large amounts of noisy data. Some practitioners suggest that subject matter expertise will be much less important in decision-making than learning from real-world big data (which is considered empirical). Data visualizations are especially important for "big data" in terms of coherence in order to spark understandings and hypotheses, to identify potential trends, to identify critical "nodes" or clusters, to identify anomalies, and so on.
To engage big data, people will rely on data visualizations to understand complex inter-relationships.
Thanks!
Thanks to Drs. Julie Rorabaugh and Jonathan Bacon for the idea for this presentation. Per their request, I hope I was able to keep this sufficiently simple...while keeping this somewhat engaging
.
|
Contact Information Dr. Shalin Hai-Jew Instructional Designer Information Technology Assistance Center (iTAC) 212 Hale Library Kansas State University shalin@k-state.edu 785-532-5262
|
