Submit Comment

show all (2)
There are no comments. Click the text to your left to make a new comment.
admin

Michael Gurnstein’s post here makes me think about how the point from Scott (1998) might be developed further – pointing to the different points at which data effects change. Gurnstein cites the digitisation of land-records in Bangalore and it’s impact on land-ownership.

reply

As a non-techy adn possibly a technophobe, these two diagrams are extremely clear and helpful. I am struck especially how leaders and managers push out information especially in times of change and instability and take little account of supporting knowledge creation, let alone wisdom! Of course the pyramid diagram suggests only a bottom up process, but i guess this is actually a dialectic: the lived lives of those on the margins may be known and understood but not recorded officially

reply
1 0

It is not uncommon to find the terminology of ‘opening access to information’ and ‘opening data’ used interchangeably (Parycek Sachs 2010). Although integrally connected, the two concepts should be kept distinct. Data can be literally defined as “a thing given or granted; something known or assumed as fact” (OED 2000). Data may be raw/primary data (the direct product of measurement), derivative data (e.g. cross-tabulations; sums; reshaped data), meta-data (data about data) or operational data (data about data use) (Floridi 2004). Whilst data underlies many acts of human decision-making, we generally operate on the basis of information, which, in the philosophical General Definition of Information (GDI) is defined as “data + meaning” (Floridi 2004). In knowledge management the data-information relationship is often expressed as a hierarchy (Fig. 2), with data as the foundation layer, and information atop it (Rowley 2007). The GDI summarizes the relationship between data and information with the thesis: “no information without data representation” (Floridi 2004), that is, all information contains data, and data is represented to become information. Generally representation reduces the amount of data to be taken in (e.g. ordering entries in a table so readers need only look at particular sections, or summarizing aggregates through visualizations or text), whilst also giving context to the data (Rowley 2007; Chi 2000). However, visualization can also increase the information that can be communicated in a small amount of space and time, as is achieved in Hans Rosling’s GapMinder visualizations of poverty statistics  (Card 2009; Rosling 2007; Rosling 2006).

2 0

Whilst in pre-Internet times, carrying out analysis data reduction inside governmental bureaucracy may have been necessary to produce something concise enough for dissemination, increased data storage capacity and processing power, standard data formats, and the development of digital data exchange (particularly online), mean that possibilities for on-demand data analysis have grown dramatically (Butler 2007). Information that was formerly expensive and complex to render can now be extracted from datasets using freely available online query and visualization tools (Viégas et al. 2007). This technological shift provides the foundation for ‘democratization’ of information creation, breaking governmental monopolies on representing an interpreting data, and fuelling calls for policy change. Mayo and Steinberg’s Power of Information Report (2007, p.27) argues for separation of data, analysis and presentation ‘layers’ of government information, opening data and analysis for re-use (Fig. 2). Rosling is clear, however, that increased availability of data and data-processing tools does not remove the need for statistical and data analysis expertise (2010).

3 1

4 1

Increased utility of raw-data can increase pressures to collect raw-data. Allan (2009) notes that much PSI is “simply not collected in a usable form at present”, either as paper records with “no electronic or otherwise easily accessible version” or in arcane and unusable formats. Snowdon (2010) has argued for a focus on the rendering of information as data, as much as a focus on rendering data as information. However, although extracting data from information can increase the re-use potential of the content concerned, the codification of information into data is not a neutral act (Bowker 2000), involving decisions that affect the ease of different future data uses. The very act of government data collection, by virtue of government’s central role in society, can also effect as well as detect change on the ground (Scott 1998).