Tuesday, August 05, 2014

Information quality

As I think I have written before, my mood seems to be very dependent of how well I'm doing with my research. I have to admit that this high level of coupling is juvenile, but I don't seem to be able to change this mode of behaviour. Yesterday, I was very frustrated with the requests of the research proposal reviewer who required explanation of how the research will have a business/management dimension, as it seemed that I couldn't find anything relevant. Today, though, is a new day and things have changed somewhat.

I am aware of two methods of finding relevant papers: the classical and the modern. In the classical method, one finds an interesting paper, reads it, notes the references then follows the references (when I was an undergraduate in the 1970s, the papers were found via a card index). In the modern method, one access a site such as Google Scholar then performs a search with carefully chosen parameters. It turns out that I'm not very good at using the modern method, as the really interesting papers have all been found by tracing the references and not by searching Google. That said, the best method of all seems to be a hybrid: finding something interesting via Google, reading it then following the references. But as references obviously can only lead to papers which had already been published when the interesting article was published, this method only leads to the past. The twist is then using Google Scholar to find the papers which have cited the referenced paper: this method leads to the future.

The first important find was a paper entitled "Organizational impact of system quality, information quality, and service quality". This paper is full of management speak such as "The value of Information Systems (IS) can be realized by improving profit margins for the firm, providing easy-to-use and useful applications, and designing easily maintainable software. IS quality as conformance denotes designing systems that conform to the end users’ information requirements and adhere to industry standards. Meeting customer expectations of IS quality is accomplished by offering appealing, user-friendly interfaces, entertaining user requests for changes, and satisfying the stakeholders of the IS. Organizational impact represents the firm-level benefits received by an organization because of IS applications. The organizational impact of IT is realized through business performance which leads to business value. Organizational impact has been measured as competitive advantage and strategic value, market value, organizational efficiency and effectiveness, and capacity utilization. IT resources create economic value by increasing operational efficiencies and creating competitive advantage." I sincerely hope that this is the kind of material which the research committee require.

The second interesting find came from a paper published in 2005 which defines four dimensions for information quality:
  • accuracy
  • completeness
  • currency (how up to date the data is, not whether it is stored in shekels or dollars)
  • format
I then remarked that "whilst the importance of the first three dimensions of information quality are fairly clear, I think that the importance of format is less well appreciated, even though it can be understood on a visceral level. The canonical paper (796 citations) on this subject (Vessey, 1991) dates from a pre-Windows 3 era, when easily programmable graphical applications had yet to appear. At the time, there was a clear divide between ERP systems and shadow systems, with no computer based interface between them. Thus by necessity, any data stored in a shadow system would have to be entered manually. 

I have long been convinced that the method in which information is presented (tabular vs graphic) has a major effect on the way in which the information is perceived; thus presenting the same data in a 'boring' table or in a 'state of the art' graphics application complete with gratuitous decoration would elicit completely different reactions in those viewing the presentation. Whilst Priority can nominally display data pre-formatted in a specific manner as a graph, such capability is very hard to achieve. Thus those wishing to display ERP data graphically are left with virtually no option other than to export the data to Excel and display from there. This subject could be added to the questionnaire and as a possible dependent variable, although I suspect that it would not be applicable to the majority of users."

Of course, citing a paper from 1991 is almost self defeating. But I then used the 'cited by' method to find more recent papers and found one published in 2014 which seems partially applicable. Unfortunately, I have only been able to access the paper's abstract but I am hoping to obtain the complete paper shortly. If this paper does not deliver any suitable quotes then I will follow its reference trail. 

I am now in a much better mood. I hope that my mentor will not deflate this mood.

No comments: