Setting Up Education-Based “Crosswalk Analyses” on an Online Survey Platform

Introduction

“Crosswalk analyses” are a fairly modern analytics approach which maps the granular elements of one system with those of another. One common approach is to map one database schema to that of another, so the data from both databases can be melded and somewhat interchanged. Outside of database administration, though, crosswalk analyses are used in a more loosely coupled way to map one system to another laterally (in one direction), and these may include schemas, ontologies, taxonomies, frameworks, standards, compliance criteria, performance criteria, datasets (including structured and semi-structured data), and others.

A “crosswalk analysis” bridges between at least two systems to identify commonalities and differences (comparisons and contrasts) between them, in order to enable integration of datasets, programmatic gaps analyses, cross-disciplinary work, and other applications. The systems compared include a range of content types, including schemas, ontologies, taxonomies, frameworks, standards, compliance criteria, performance criteria, datasets, and others. Crosswalk analyses are considered efficient because they maintain the integrity of the compared systems—by keeping the terminology and phrasing verbatim—by showing connectivity at a basic unit level of analysis for each respective system. As such, they do not break down silos but connect “content silos” (Johnston, June 22, 2015). [Note: Such crosswalk analysis bridging can also occur more abstractly or at a higher level of abstraction, but for usability, the precision at the most granular units of analysis seem to be preferable. These have also been referred to as “equivalent elements” (“Schema crosswalk,” Sept. 28, 2018).] The innovation of this analytic technique is in the crosswalk, and the overlap between the systems is somewhat interpretive (and defined by the objectives of the crosswalk analysis). The crosswalk itself is partial and selective and does not include all potential overlaps between the two disparate systems. (Figure 1) Or, not all overlaps between two systems will likely be seen as relevant. This is not to suggest that some crosswalks may not be comprehensive, and in some cases, that level of detail may be required and often depicted in a crosswalk analysis matrix. (The essential structure of this is to have one system represented in the row headers down the leftmost column, and the other system represented in the column headers across the top row.)

Figure 1.

A Venn diagram analogy to the crosswalk analysis

In Figure 2, the bridging or connective function is depicted. The systems that are being studied are on each side of the crosswalk, but the main focus is to see how the left system crosswalks to the one on the right, with the main focus on the original system (shown to the left). How does Metadata Schema B map to Metadata Schema A? How do professional competencies map to the professional standards? How does Framework B map to Framework A? How does Dataset B map to Dataset A? The question is how the second system maps to the first. The idea of a focal system matters because of the “lateral” or one-directional nature of crosswalks in many cases. (Figure 2) More on how crosswalk analyses work will be presented in the body of the paper.

Figure 2.

A crosswalk analysis (with bridging or connective functions)

Since crosswalks are usually built by subject matter experts or content experts with expertise in particular disciplines, it is important to be able to access expertise across geographical distances. While there are benefits to using co-edited documents hosted online (whether as first-party or third-party platforms), online survey systems offer a range of benefits to creating crosswalk analyses. This work describes some basics about building such an instrument on a modern-day online survey platform.