In order to gain some familiarity with Web Grid and Kelly's Personal Construct Psychology, I endeavored to develop a Web Grid of my own. I chose to create this grid based on some of my favorite hikes in Banff National Park and the Kananaskis. The elements in my grid consisted of the hikes themselves and I chose six of them. The constructs represented the criteria used to evaluate and rate the hikes. I included such things as accessibility, scenery, elevation gain, and traffic levels, to name of few.
Initially, I found the Web Grid somewhat difficult to figure out. The terms "constructs" and "elements" were quite ambiguous and I had to look at a few examples and read portions of Comparing Constructions Through the Web to get me started. Once I determined how the system worked, it became to relatively easy to construct my grid and I was able to see the usefulness and practical applications of the WebGrid system.
I found that it was helpful to be consistent in deciding on the rating scale. By this I mean that if my first construct was "boring vs. great scenery" and I rated "boring" as 1 and "great scenery" as a 5, then it was necessary for me to continue to rate desirable constructs with higher values and less desirable ones with lower values. When I followed this guideline, the graphs obtained from PrinComand FOCUS displayed results that seemed to be more accurate.
It was interesting to see another perspective of the results of my grid using PrinCom and Focus. However, it was necessary to reload the image every time I wanted to see my grid from a new or different perspective and this proved to be slightly annoying.
The easiest way to develop constructs was to use the triad elicitation method. The system presented me with three elements and then prompted me to uncover a way in which one of the elements differed from the other two. This significantly simplified the problem of finding unique constructs among the 10 elements. For instance, I was initially presented with the three tools: Electronic Publishing, Object-Oriented Systems (OOS), and Cross-Platform GUI's (PIGUI). The obvious difference among these three is that almost everyone is involved with Electronic Publishing, whereas OOS and PIGUI's are used mainly by computer programmers. I continued using triad elicitation until I had come up with six unique constructs.
The constructs I developed were remarkably similar to the constructs developed by Gaines, in terms of their distinction. The major difference was in our respective uses of terminology. According to the four-way classification structure, our two grids show a high degree of correspondence. This proved to be the case for each of the constructs I developed. That is, for every construct I developed, there was a construct of Gaines' that represented the same concept but was described using different terminology.
The most striking example is my construct "Stored information vs. Networked information". Gaines' grid contains the construct "Application technology vs. Communication technology". The two constructs convey the same meaning but use different terminology.
I created this grid based on the presentation I gave on data encryption. The context of my grid was "Secure Encryption Methodologies and Products". I included the public and secret key encryption methodologies and encryption products like Pretty Good Privacy and anonymous remailers as my elements. My constructs included security, usage level, and complexity, among others. The grid made it easy to spot secure and insecure encryption products and methodologies.
The PrinCom utility distributed the elements among four quadrants. It did seem to produce some minor inconsistencies. The graph indicates that public and key escrow encryption differ widely in terms of the security construct. However, I gave these two elements identical ratings on this scale. Also, a person analyzing the graph would be lead to believe that key-escrow encryption is a very popular encryption methodology. In fact, it is quite the opposite but this is not accurately reflected in the graph. The graph does, however, model the Clipper Chip exactly as I had intended. It is depicted in the graph as a highly controversial, rarely-used encryption product.
The FOCUS facility is, in my estimation, a much more practical and useful tool. It consistently does an accurate job of grouping similar constructs. It is fast and easy to correctly determine the attributes of the individual encryption products and methodologies.
In summary, it easy to see the value of a system like Web Grid in practical applications. On the whole, it does a good job of grouping relevant elements and of making differences apparent. It does have, however, a couple of minor glitches such as the need to reload images to view new perspectives.
Comments? Send e-mail to Sandra McCallum