Page 100 - Proceedings of the 2017 ITU Kaleidoscope

2017 ITU Kaleidoscope Academic Conference

Table 1. Influence factor clusters and means
Key Description Mean
Intrinsic value based 5.64
X9_confidential The confidentiality of the information 5.66
X9_secret The secrecy of the information 5.11
X9_privacy The privacy of the information 6.16
Intrinsic personal 4.7
X9_benefit That the information is used for my 5
direct benefit and wellbeing
X9_control The need to control access to personal 5.26
information
X9_alone My right to be left alone 3.84
Extrinsic 4.48
X9_UN Assisting the UN to compare quality of 4
health across different countries
X9_reporting Assisting government to report on 4.34
quality of health in the country
X9_industry Contributing to improving the health 4.84
industry through better medicines
X9_national That the information is used for 4.74
national health policies

On average the factors that the participants point to as
Fig. 6. Correlation matrix of influence factors having the highest influence on their willingness to share
their personal health information are those associated with
The factor that scored the highest (mean 6.16) on the privacy, confidentiality and secrecy aspects of the
influencing individuals to share (or not to share) their personal data. These three concepts are of course tightly
personal health information is the privacy of the coupled and represent varied conceptualizations of the
information, with the lowest being the individuals’ right to notion of privacy [34]. The factors associated with the
be left alone. A subsequent clustering of these factors was external use (e.g. contribution to the national or UN-level
undertaken, using hierarchical clustering based on the social indicators efforts) of the individuals data in general
Euclidian distance off the Spearman correlation matrix of are of lower influence on the participants towards them
the factors (Figure 6), and using the complete linkages sharing their data.
agglomeration method to find similar factor clusters (
Figure 7). The two broad findings, on the clustered differentiation of
the entities within the data ecosystem that the participants
are happy to share their data with, as well as on the clusters
of factors that influence individuals to collect and share
their personal data, give further insights on the strategies
for engagement of individuals within the sustainable
development data ecosystem.

4. DISCUSSION

Mortier el al have suggested that legibility, agency and
negotiability are three of the core themes for consideration
in Human Data Interaction [17]. These three themes are
associated with amplifying the capability of individuals to
engage meaningfully and actively with their data. This
research has engaged with these concepts from the
Fig. 7. Hierarchical clustering of the factors
perspective of small data to support the individuals’
participation within the sustainable development data
Three key categories emanate from the clustering exercise ecosystem. The formulation of small data for development
undertaken, and these have been labeled: 1) the intrinsic in section 2, which is framed around the Capabilities
value based factors – which reflects internally held Approach, emphasizes individuals’ freedom and agency to
individuals beliefs about the values and attributes of the use data to achieve the development ends that they desire.
data; 2) intrinsic personal factors – which are factors Recognition is given to the fact that data, including small
centered around the individual and their interaction with data, has the potential both to restrict and to expand
their data; and 3) extrinsic factors – these are factors that individuals’ substantive freedoms towards their
are associated with externally-oriented use of the development and wellbeing, as such the individuals
st
individuals’ data. Table 1 lists these categories and the participation and engagement within the 21 century data-
associated mean scores for the factors and across the driven societies needs to be critically considered from this
categories.

– 84 –