A few things we learned thanks to the Issue Safari

On December 12 2012, the EMAPS project organized its first Issue Safari.
Such workshop was an improved version of the seminar we organized in June 2012. In both cases the objective was to submit some of a set of maps we had prepared on the theme of aging in UK to the evaluation of a selected group of potential users.
The first workshop had been very interesting, but had also revealed the great difficulty of finding a common ground between mappers and users. Users were very critical on several aspects: from the choice of the research questions, to the datased employed, from the visualisations employed to the legibility of the maps (see emapsproject.com/blog/archives/1329).
In order to overcome this difficulty, we started a new process of interaction with the users (described here emapsproject.com/blog/archives/1701, here emapsproject.com/blog/archives/1728 and here emapsproject.com/blog/archives/1754). This process helped us to produce a more effective set of maps and to identify a groups of users potentially more interested in our visualisations. The December Issue Safari was the result of this process.

I think it is fair to say that the Issue Safari has been a success in the sense that we have greatly improved the adaptation between users and maps. Most users found most maps relevant and engaged with them in interesting ways. This allowed us to collect a richer user-feedback that we will now be able to re-invest in the case study of climate change adaptation.

Drawing on a very detailed set of notes collected during the Safari by the facilitator of the Young Foundation and compiled by Lucy Kimbell, I have prepared the following synthesis of the lessons we learnt during this workshop

General Observations

On the bright side

  • The description of the protocol used to collect the data and design the maps improves both the engagement and the understanding of the maps.
  • User appreciate having different visualisations of the same data and are generally comfortable in navigating from one to the other. The connection between these visualisation however has to be made very clear.

To improve

  • The possibility to observe the temporal dynamics/trends would have improved almost every map.
  • Users would profit from some indications or examples on how to use the maps.
  • Users asked both for more details and for more aggregation. They wanted to be able to zoom in and out the maps.
  • The possibility to interact with the maps should allow users to:
    • Filter (hide part of the information)
    • Rank (make part of the information more visible)
    • Aggregate (go from data points to categories)
    • Search and focus (observe specific data points and their neighborhood)
  • Maps are considered irrelevant/uninteresting if:
    • They are too general/superficial
    • Too disordered/untidy

General Observations

  • Comparison (of different country, actors, moment in time…) is the most common strategy to make sense of the maps.
  • Previous knowledge of the topic determines what users will look for in maps
    • Non-experts on the topic are interested in understanding which is the norm.
    • Experts on the topic are interested in identifying the deviation from the norm.
  • Previous knowledge of the topic determines the touchstone for evaluating the maps:
    • Non-experts want maps to be ‘interesting’ (do I learn something from the map?)
    • Experts want maps to be ‘useful’ (how can I use the map?)
  • (Non-exhaustive) list of possible maps functions:
    • To get an overview of a topic (understanding the landscape
    • To prompt me to look at things
    • To confirm what I already know
    • To find support to make my argument
    • To get inspiration and find new sources of information
    • To expand my current understanding
    • To find allies and resources to connect to
    • To get access to resources I don’t usually use
    • To identify dynamic sites and opportunities to intervene into
    • To see how discourses change over time
    • To reveal how particular communities’ discourses vary
    • To prompt reflection on own discourse
    • To prompt reflection on how I conceptualize my practice/network
    • To access an edited, trusted reading list that is regularly updated

Specifics maps

3A – Cross-national comparison of aging-related issues

  • The cross-national comparison were appreciated but the maps was perceive as too generic/ superficial.
  • The protocol was not explained clearly enough.
  • It was difficult to make sense of the lists – some aggregation would have been useful (positive/negative spin of the issues – types of issue).
  • Comparison over time would be much appreciated.

0A – Web cartography of the most important actors on Aging (from the UK viewpoint)

  • Remains one of the most appreciated maps (helpful to navigate a new topic and identify the key players on an issue).
  • The direction of the link should be kept.
  • The meaning of the position of the nodes should be explained.
  • The possibility to search a specific node would be very useful.
  • Comparison over time would be much appreciated.


2A set – Occurrence network of expression and documents

  • The mapping helps understanding
    • who is setting the agenda/language on aging (center of the graph)
    • who deviates from the normal language (margins of the graphs)
    • who share the same language (clusters)
  • Expressions should be categorized by theme
  • The map is ‘untidy’ and difficult to read (edges are too visible, expressions not enough)
  • Comparison over time would be much appreciated.

2B set – Co-occurrence analysis of words in different set of documents

  • These maps are especially useful for non-expert users.
  • The synergy between networks and list is very effective.
  • Comparisons among the maps extracted from the different corpus is the most interesting function (but the miniatures were too little to be useful).
  • In the networks it is difficult to identify clusters of words. The networks tend to be used a mere word cloud.
  • Expressions should be categorized by theme.
  • Could be used together with the previous set:
    • 2A to explore the proximity in the employed language
    • 2B to identify the specific words used

1B – Heatmap of the most controversial element in a Wikipedia page
 

  • “The most interesting and yet the most difficult map to use”
  • It would be important to know not only which elements are debated, but also why.
  • This type of analysis is more interesting for scientific-related subjects
  • The interest of the map depends on the actual visibility/traffic of the page (do people go to Wikipedia to find information on aging?)


1A
– Network of dementia-related blogs and the online resources they cite

 

  • The lack of categorisation of the blogs is perceived as a major problem
  • The bubble chart is the wrong type of diagram for visualising the data.
    • it makes it difficult to compare the size of categories and time
  • It would be interesting to compare the hyperlinking and the semantic linking
  • The ‘artificial’ structure created by the linking strategy of Robert De Marco should be flag up to users.

 

Leave a Reply

Spam protection by WP Captcha-Free