Special Interest Group: Oil and Gas

Dear colleagues

Attached are some recent published research findings you may find of interest. They have applicability outside of the oil and gas industry, so feel free to cross post to other forums (I am unable to do).

Thanks, Paul

Discovering high value information in the enterprise : No link between self reported search expertise, user satisfaction and actual performance

A study of experienced information professionals in a large multinational oil and gas company using the company enterprise search/library tools, indicated when 'under information overload' (typical in enterprise environments) there was no relationship between 'user satisfaction' and how well each participant had 'performed the search task'. The study focused on exploratory search (gathering many information items). Furthermore, there was also no relationship between self reported 'search expertise' and how well each participant 'performed the search task'. The results were fed back to the searchers and senior management, both groups were somewhat surprised at the poor performance. The implication is that many large organizations may be oblivious to the 'search literacy' of their staff and capability in this area. The study provided evidence of crucial information being missed in the workplace, supporting existing literature where poor search has been responsible for missing evidence of fraud and has caused fatalities in the health sector. Organizations may not be applying a 'systems thinking', organizational learning approach to search capability, with an over emphasis on IT. Interviews with Knowledge Management and IT staff in smaller oil and gas companies and organizations from the Governmental (Defence), Aerospace and Pharmaceutical sectors indicate some of the findings may be transferable.

More here: www.paulhcleverley.com

------------------------------
Paul Cleverley
Robert Gordon University
------------------------------

Hi Paul,

Can you expand a bit on "there was also no relationship between self reported 'search expertise' and how well each participant 'performed the search task'". Does this mean the users over estimated their search expertise? Or was their expertise not relevant to conduct a "good" search?

Best regards,

Dennie

Hi Dennie

There is an extended abstract and link to the full paper in the link I added to the post www.paulhcleverley.com (most companies should have a Wiley subscription so they can download, otherwise please contact me if appropriate).

The users rated their search expertise using a Likert Item (1=very poor, 2=poor, 3=neutral, 4=good, 5=very good). Of the 26 participants, 85% rated themselves as 'good' or 'very good'. When analyzing (using a Spearman Rank test), there was no statistically significant association between self reported search expertise and how well the task was actually performed. A conclusion is that in general participants did overestimate their search expertise.

These were experienced IM staff, yet they only found 27% on average of the high value items in the experiment. Why has this happened? One inference is that people 'are not learning' from each other - social learning (or from formal organizational learning) when it comes to exploratory search.

Hope that clarifies

Regards, Paul 

Hi Paul

It's an interesting topic. Thank you for sharing.  I guess promoting activities like tagging or keywords would help address this. If for example a geologist works on a report and believes it is relevant to another study done in another location, even though that location is not specifically mentioned in the report, he or she would use they keywords field for this information to be captured. So effective sharing of tacit knowledge to ensure when a user conducts a search the best possible results are returned.

Thanks

Hello Tita

Indeed, quoting from a paragraph in the conclusion in my paper:

"Advances in machine learning, semantic networks, and enterprise information tagging strategies could make it easier to locate information in the workplace, mitigating mismatches between the queries used by staff and the information they seek. However, for a variety of reasons, such as content structure and availability of statistical data for algorithm deployment within the enterprise, it is likely that search expertise levels will play a crucial role in exploratory search task performance."

Getting a little philosophical, I have observed that RM/IM practitioners (myself included) can have a tendency to focus on the technical (IT) and formal aspects of information management, sometimes forgetting an organization is effectively a social system. This social system is a Complex Adaptive System (CAS); we are sometimes surprised when things in IM do not play out 'as designed or intended' but perhaps that is inevitable. Tagging documents is a good example, where what is written down or articulated as a practice in organizations, and they way people actually work have always been quite different things.

Maybe if we adopt a mind-set that ignorance, fallibility and error in Information Management 'are inevitable', we may adopt different approaches to both managing information and searching for it.

Cheers, Paul