DANA Software

DANA is a product of academic research, aimed at improving the professional practice of policy analysts. My interests are academic, not commercial. This means that I like to learn more about actor network analysis from people who use DANA, and obviously it also means that I like to get academic credit for my work on DANA. Therefore, you are free to use the DANA software, provided that you refer to me and my university whenever you present or publish results obtained through this use. If you earn money by doing actor network analysis, that's great. But you are not allowed to sell the DANA software, nor even to distribute copies to others at no charge, because this impedes my research of its use. If you want to share DANA with others, refer them to this web site and let them get their own copy. These conditions are stated in the license agreement that is displayed when DANA starts up.

Obtaining DANA

Academics, students and professionals are encouraged to become active members of the DANA community. To obtain the DANA software, you must provide some personal information. I will not use this information for commercial purposes. I simply like to keep track of who may be using DANA, as this will allow me to conduct a survey at some future point in time.

NOTE:  Unfortunately, many virus scanners block messages containing executable files, even when zipped. The DANA software is therefore no longer mailed directly to you.

Upon completion of the form below, a message with a link that allows you to download the DANA software will be sent to you automatically by e-mail.

Function:    at a  

Version history

DANA is not a finished software product, but a "live" object of research. Depending on the time I can devote on its development, new versions of the DANA software will be released. The version number is displayed in the caption of the main window and consists of three numbers separated by dots, e.g., 1.0.2. The first number indicates the major version, the second number the minor version, and the third the release number. The release number will increase after minor changes (typically bug fixes) that do not add new functions to DANA. The minor version number will increase when new functions have been added and/or the user interface has been changed. As for the major version number, I do not expect this to change in the coming years...

DANA version list

NumberRelease date Description
1.3.32010-02-01 00:00:00Some bug fixes:
  1. Conflict analysis: Since version 1.3.0, DANA calculated the utility of a single tactic on the basis of ALL goals. By consequence, this utility included the utility of "no change" in the goal factors that are not affected by the tactic. This often led to overinflated conflict scores.
    In this new version 1.3.3, DANA only adds up the utility over the goals to which there exists a path from the evaluated tactic.
    NOTE: When evaluating a single tactic, DANA also takes into account changes caused by prospects.
  2. Issue linkage analysis: One minor bug (# conflicting interests was represented as a percentage) and one bigger one (percentages were computed not relative to the count in the row arena, but to the total number of factors in both arenas) have been fixed.
    NOTE: DANA calculates same/conflicting interests by looking only at goals and actions, testing in each arena which change direction (increase, no change, or decrease) is generally preferred (i.e., produces the highest positive utility for all actors in the arena -- conflicting interests within one arena are not counted in this analysis).
1.3.22009-02-03 00:00:00DANA finally facilitates "merging" two factors. This is especially useful for resolving synonyms and antonyms when perception graphs made by different authors need to be combined in a single case. Assume, for example, that some perception graphs contain the factors "pollution" and "energy efficiency", while other perception graphs contain the factors "polluting emissions" and "energy loss", and that the analyst considers "pollution" and "polluting emissions" to represent the same aspect (synonyms) and "energy efficiency" and "energy loss" also (but as antonyms). DANA can now automatically replace all occurrences of a factor F by a factor G while preserving the consistency of goals, prospects and causal links. Look here for a more detailed explanation.
1.3.12008-10-01 00:00:00Added features:
  • Possibility to export graphs as Pajek files.
  • Possibility to indicate whether a factor represents a "stock" or a "rate" (System Dynamics terminology). The idea is that when simulating consecutive rounds, changes cumulate in stocks, not in flows. This feature is very much in a testing phase still!
1.3.02008-03-02 11:00:00Major changes:
  • DANA finally complies fully with Bayesian logic. This has a price: when many uncertain links and/or prospects are defined, evaluation of a strategy becomes quite slow.
  • More "natural" default utility vectors for goals.
  • By consequence, you may obtain different results when running analyses on old cases using the new DANA. See also the new features below!
Powerful new features:
  • Custom goals: In the goal dialog you can now set any "face" for each of the 7 change levels by clicking on the face. LEFT mouse button increases the utility value (happier face), RIGHT mouse button decreases it (sadder face).
  • Aggregated perceptions: Until now, DANA could not infer strategies for aggregated actors. Now it can! Label actors with tags and then use the Grouping dialog to define an actor group (e.g., all actors with a green tag) and select "aggregate by group". When you run the "Inferred strategies" analysis, DANA will first infer the "group perception" by merging the perceptions (including goals) of the "green" actors into a single perception graph, and then search for the strategy that maximises utility for the group of actors a a whole.
  • You can tell DANA how to combine influences. When several links go into a factor, DANA used to always combine these using a (Bayesian) + operator. By default, this is still the case, but you now have the option to specify for a factor how it changes as a function of its "incoming" changes: ADD (= default), MULtiply, or take the MINimum or MAXimum of the incoming changes.
1.2.12007-10-18 18:00:00

Bug fix: Error messages popping up during Multi-Criteria Analysis

Minor new feature: Easier way to rename factors. No need to switch to dictionary mode; instead, select a factor in a perception graph and press Ctrl-R (or select menu item Edit|Rename). The Rename-dialogue also allows you to change factors into attributes or actions and vice versa!

1.2.02007-09-13 00:00:00New feature: actor relations (affect, co-operation, dominance, frankness) can be specified. Consequences of such relations for preferred strategies to be implemented, still, but coming up soon!
Lots of improvements, though:
  • better compliance with Bayesian logic: when factors are affected by multiple causal paths that share uncertain links, the uncertainty of these links no longes seen as "independent stochastic events"
  • better memory management: DANA used to slowly but surely eat up memory resources; now most of the memory leaks have been fixed
  • more transparent commenting: DANA now shows which object (actor, factor, causal link, ...) you are adding a comment to in the status bar at the bottom of the window
  • interruptable analyses: just click on the newly added "cancel" button in the analyses toolbar
  • graphic output of actor and arena level analyses: see similarity of perceptions, conflict, and resource dependency visualised as arrows between actors
  • DANA now also saves the last analysis and group configuration settings
  • more small things I forget...
1.1.12007-07-27 00:00:00Bug fix: goal adaptation after simulating one round of actions.
1.1.02006-10-29 11:30:00

Bug fixes as well as new features, some of which may (of course) contain new bugs...

  • Some calculations in analyses (notably Conflict) were counterintuitive and have been improved.
  • As analysts show a tendency to draw causal links from one action to another (to express "when A does X, then B will do Y"), this is now dealt with correctly when inferring strategies: actions with incoming arrows are no longer seen as "deliberate actions" and will appear as "no tactic" in inferred strategies. The same holds for actions that have been marked as prospects.
  • A truly new feature (that justifies moving to version 1.1) is that DANA now supports a crude type of simulation, using the Analyst view as "reality" and creating new arenas to show actions and the resulting "state of reality" in subsequent rounds. This new feature will be documented some time in the future...
1.0.52004-07-03 18:04:20Bug fix: "Clustered by group" option in grouping for analysis should now function properly.
1.0.42004-06-09 11:36:30Modifications: added Details option for MCA; prevented line breaks at minus signs in reports
Bug fixes: rare numeric overflow during analysis; error in progress message when analyzing strategies
1.0.32004-06-05 11:00:00Bug fix: the utilities associated with standard negated goals (e.g., " should NOT increase") are now correct.
1.0.22004-05-31 21:00:00More bug fixes. Changed to sampling strategy for optimization algorithm and modified presentation of results.
1.0.12004-05-28 13:00:00Second release to beta-testers. Several bug fixes and some actor level properties implemented.
1.0.02004-05-26 21:00:00First release with only factor level properties implemented.

Last update 2007-05-18