GIST: A model for Design and Management of Content and Interactivity of Customer-Centric Web Sites—By Albert et al.

The methodology and framework proposed makes a great effort at trying to identify those unknown users.  In some cases, all of the information gathered does not go further than identifying who the user’s ISP is?  And in others, a little more is collected.  The GIST (Gather-Infer-Segment-Track) methodology seems to require much synchronization among the various resources.  However, in most of the cases the users remain a mystery, but their behavior with the web site tend to provide useful information in order to improve the representation of the site and the hit rate.

Albert et al. relied on the design science research guidelines presented in Hevner et al. (2004) (Design as an Artifact, Problem Relevance, Research Rigor, Design as Search Process, Research Contribution, and Research Communication).  I am amazed at how fast Albert et al. (June, 2004) produced and published their GIST methodology paper, given that the Hevner et al. paper was published just a couple of months prior (March, 2004).

Although it seems that employing GIST would require significant resources, but then again the sort of businesses that just need web presence (non-transactional based) tend to be large companies with many resources available at their disposal. 

This paper was published in 2004, it seems that technologies of identifying customers behaviors were available much earlier (late 1990s).   Nonetheless, the paper served as a good example of designing an artifact under the guidelines of design science research.

Hevner et al argued skillfully for the utility (usefulness) and value of design science research.  Their main goal was to sell us (researchers and practitioners) on design science research.  I am sold! I found their presentation to be very useful in aiding us with the means of how to conduct design research.  The design science research guidelines were particularly helpful.  They provide clarification for what needs to be done. 

I was not completely clear on the issue of rigor versus relevance as they presented it.  They seemed to state that more rigor tends to take a way from relevance (please correct me if I am wrong).

As I was reading, I had a question as to whether design can be considered research.  I found my answer here: http://www.isworld.org/Researchdesign/drisISworld.htm.  I also wondered about how different is design research from action research.  There appears to be many similarities between the two types of research.  I am growing to value this kind of research more and more, because I feel that producing an artifact is the pudding of the research.   

Also, I was able to relate more to the reading of this article, thanks to the hard work of CGU faculty for organizing the first design research conference DESRIST 2006, here at Claremont.  At the conference, I met Alan and many other scholars.  I look forward to DESRIST 2007 on May 13^th, 2007.

Rube and Pare’s intended target for this article appear to be future researchers and possibly reviewers and editors.  It is quiet surprising that for a positivist case of research and inquiry that they only found 42% to have posed a research question.  These were papers that were published in the top-tier journals.  Furthermore, only a meager 28% elucidated their data analysis process. 

In essence, they appear to be proposing a set of guidelines that positivists case study research should include.  While they also caution against using the attributes that were highlighted in Table 3 of their finding, since those attribute do not guarantee rigor in future research.  Nonetheless, they do form a template even without the guarantee of rigor.  Future researchers should find this as a useful tool to guide their research.  They also should understand that only rigorous work will pull through.  When comparing this to Reynolds’ cartoon selection of “Dennis The Menace” (p.148), it is Margaret’s work of fitting the pieces of the puzzle together that reflects rigor.  Dennis’ way certainly lacked the rigor, the pieces were forced to fit together, and this made the result unclear about what that puzzle depicted.

I suspect that deciding on which strategy to choose for developing a scientific body of knowledge might have played a  part in the case research.  Whether research followed either “research-then-theory” or “theory-then-research,” or somehow a hybrid of the two might have lead to some confusions. 

Also, it was not until 1994 that Yin made his contribution of knowledge to case research, but the study employed his contribution as a gauge for earlier studies.

The authors presented their claims according to a set of hypotheses in a 2X2X2 covering Task Complexity (Low vs. High), Query Interface (Text-based vs. Visual), and Spatial Ability (Low Vs. High) and their impact on three dependent variables (Subjective Mental Workload, Decision Accuracy, and Decision Time) (Figure 1. The Research Model).  The research question: “How do information visualization techniques such as visual query interfaces influence decision making performance?”   The authors clearly stated their objective and skillfully narrowed their focus of the interplay between text-based and visual interfaces to the individual’s spatial ability.

First, When I read the abstract, I was glad to see that they also included the results of their study in the abstract.  However, I had to dig deeper to find the basis and their theoretical framework.  I also thought the Media Richness Theory, first proposed by Daft and Lengel (1984), could have been relevant to explain some of their findings.  But, when I looked closely at it, I found out that it was not specific to decision making, yet it is relevant to the task characteristics and task performance. 

Second, the authors focused strictly on spatial ability, despite having text-based tasks.  I also thought that linguistic ability could have been highly relevant, but this could also the direction of a future research, as it may increase the scope and focus. 

Third, I found Table 1, Summary of Hypothesis Testing to be very helpful in presenting the variables, both test statistics (F ratios and p-values), effects, means and standard deviations, as well as whether the individual hypotheses supported or not.  This provided a good summary in one place.

The findings of the study served as the evidence that provided support for the claim.  Issues that revolve around the evidence seem to have addressed.  The results are deemed reliable, accurate, and sufficient.  Although the study was conducted on a convenient sample of students, a question may arise as to whether the sample was representative.  Overall, I thought the paper was well written.  They presented several opportunities for future research and direction, which also may have been indicative of some of the paper’s limitations.