Converting diagrams to matrices : Using matrix multiplication for path analysis. In this project, I used a multi-layer network analysis, multiplying two matrices (possible paths for templates and actual paths from exit pages) to obtain unique user paths. These were then aggregated for system stats purposes. This model uses equivalence classes to partition users based on the skipped steps in their user path. I used this for mapping conversion funnels but this same modeling could be applied to a variety of case, as long as they meet the following pre-requisites: (1) the process is sequential, (2) has multiple possible paths and (3) has variable constraints or conditions. Example of such other cases could include: tracking supply chain routes, modeling patient treatment paths, planning project workflows or approval processes, planning resource allocations, etc).

This episode covers how to let users define their own aggregate criteria. This particular project covered dashboards but this can be applied to other types of projects as well.

Today we’re covering a real case study with insights on : (1) Naming conflicts, (2) encoding ambiguity (also called data representation conflict), (3)inconsistent encoding (also called non-deterministic representation) and (4) dual encoring.

When is it more advantageous to consume a more complex data chart? Decision-making considerations drive this important choice. In this episode, we illustrate the process through which a manager can estimate the complexity of his request, based on the data found by the data scientist or data analyst.

Reflections on data sets where the underlying characteristics of "the typical record" might be harder to identify.

Are you running into errors in your data analysis? Here are four quick check-points to reflect on possible error sources. Use them to flesh out your troubleshooting plan.

Being organized is a pre-requisite for analysts.  And by their very nature, analysts tend to adopt productivity systems and design their environments for efficiency.  Many decisions go into productivity systems - one of which being the analog, or digital nature of things.  An analysts analyzes... productivity systems and fads.

What makes a good data analyst?  With statistical tools and packages that increasingly do the heavy lifting for data analysts, what makes the job of the data analyst difficult?   How can you spot a competent analyst?

Debate around use cases is rampant.  Are use cases worth the time and money?  When should they be produced, or avoided?  We discuss the use case... for use cases.