547 Project III

Web Grid

Jeanette Long (long) Homepage , Mail

Comparing Constructions Through the Web
Creating a New Grid

I. Exploration

Exploring WebGrid

In the initial exploration of WebGrid, the domain that was chosen was Books, and the specific context selected was Children's Fiction. Ten written works for children were evaluated, with an attempt to include books with a wide variety of characters and storylines. Each book was rated on a scale from 1 to 5, on ten constructs ranging from the type of protagonist or hero, to the gender that the book was aimed at.

Personal Constructs on Children's Fiction
With regards to relationships between the selected works, the FOCUS grid demonstrated that "Heidi" and "Anne of Green Gables" were very much alike, scoring similar scores in many categories. Other paired books included "Little Women" and "The Secret Garden", and also "Oliver Twist" and "The Adventures of Tom Sawyer". The PrinCom grid showed little clustering of constructs, as the constructs formed a star formation. However, the display did seem to indicate that the books aimed for females generally had female heroes and human characters, and had less adventure. On the other hand, books aimed for males usually had male heros and animal characters, with lots of adventure.

Initial Impressions

WebGrid seems to be quite useful in organizing the relationships between different elements, as they are understood in the mind. It was helpful in giving concrete representations to the similarities and differences between different elements. In using WebGrid, it was found that the graphs often depicted relationships that had not been previously considered. This is likely the most advantageous and interesting feature of WebGrid. As for the grid for children's novels, it was interesting to see so many books in which the central character is an orphan of some sort or another!

One of most glaring deficiencies of the WebGrid system however, is the lack of an on-line help system. The terms elements and constructs can be somewhat confusing at first, and at times, it may be difficult to think of appropriate constructs, even with the aid of the triads. Furthermore, the tools that users can access to graphically display personal constructs have no accompanying explanatory text or comments. Of the graphs that were provided, the PrinCom graph seemed most intuitive, although all graphs would have benefited from a legend and/or accompanying instructions on how to read the graph. This would save the user the trouble of having to refer to "Comparing Constructions on the Web". Once the idiosyncrasies of the system have been mastered however, WebGrid is fairly easy to use.

On a side note, the Display option did not seem to work properly. Whenever one of the Display, PrinCom or FOCUS buttons was pressed, the graphic that was displayed was the one that had most recently been displayed, and not the graph with new constructs/elements added. To see the new graph, the Reload button had to be pressed, although this was not indicated to the user. Note that this is not problem of WebGrid per se, but may be related to the use of the Netscape Browser.

II. Elicitation of constructs on CPSC547 topics


Seven constructs for this domain were determined through the use of the triad button, and by considering some of the key aspects of each topic (eg. the graphical nature of Visual Programming and the Interactivity of Virtual reality). None of the constructs that were considered had exact Correspondence with those already given by Gaines, although there was some Consensus . For example, "Application to General Users" (myself) corresponded with "Human-Oriented tool" (Gaines). Also "Static Knowledge" (myself) corresponded with "Only act as programmed" (Gaines). Despite the difference in terms used, it was apparent by the element scores that myself and Gaines had agreement on the meaning of these properties.

Even so, for some constructs, although the elements were ranked similarly by myself and Gaines, it is debatable whether or not the same construct was meant. For example, although elements ranked similarly on the constructs "Interactive-Non-Interactive" (myself) and "Development Tool-Application" (Gaines), it is likely that these two constructs are NOT the same.

In any case, there generally did not seem to be any conflicts in the use of terms. That is, there did not appear to be instances in which the same terms were used, but different constructs were meant.

Personal Constructs on CPSC 547 topics


The PrinCom grid showed loose clustering of some of the constructs. For example, using my constructs for the domain, the grid loosely grouped together "Non-Interactive" systems with Static Knowledge". Conversely, "Dynamic Knowledge" was grouped with "Interactive" systems. The PrintCom graph also demonstrated that "Text-oriented" systems are in more advanced stages of development, whereas the state of the art for graphically oriented systems appears to be more primitive. Using the FOCUS option, "Multimedia and hypermedia" and "Information Highway" were shown to be closely related topics, as were "Electronic Publishing" and "Digital Libraries".

III. Elicitation of constructs on EDI/EFT


For the domain of Electronic Data Interchange (EDI) and Electronic Funds Transfer (EFT), various components of these technologies were used as elements. These included Value Added Networks, Digital money and Smart Cards. Some of the constructs used for this domain included those used for part II, including "State of the Art" and "level of Human Interaction".

Personal Constructs on EDI/EFT


Again, as in previous grids, the PrintCom graph was more or less scattered, although some patterns emerged that could be expected. For example, the elements "Automated Payment" and "Digital Money" fell into the same quadrant. Similarly, supportive components of EDI/EFT such as "Standards" and "Security" also fell into the same quadrant. The FOCUS graph showed similar relationships between the elements, also grouping "Smart Cards" with "Automated Payment" and "Digital Money", and "Value Added Networks" with "Security" and "Standards".

The grid for this domain was fairly difficult to construct as this topic was presented in class in fairly general terms. Furthermore, the topic is extremely broad, and it was difficult to find constructs that applied to most of the elements.

IV. Some Final Comments...

As previously stated, WebGrid seems to be quite useful in helping a user to organize his or her thoughts. Patterns between elements are likely to emerge that are not immediately obvious to the user. It is likely to be most helpful when a great deal is known about a subject, where it can be used to articulate ideas and concepts about that topic.

Even so, "Personal" Constructs are subjective. While it may be helpful in collaborative work as discussed in "Comparing Constructions on the Web", the grid is often meaningful only to the user, and not to anyone else. If the domain is too broad, it may be difficult to find any useful patterns, and if the domain is too narrow, it may be difficult fo find appropriate constructs.

Last modified March 22th, 1996
Jeanette Long (long) Homepage , Mail