NETWORKS OF PREDICTIONS

<body topmargin=0 leftmargin=0 marginheight=0 marginwidth=0> <table cellpadding=0 cellspacing=0 border=0><tr> <td valign="top" colspan=2 height=90 BGCOLOR="#336666" TEXT="#080000" background="062b85a3.jpg"> <div> <FONT SIZE="5" COLOR="#EEEEEE" FACE="Arial" ><B>NETWORKS OF POSSIBLE FUTURES</B></FONT><B>    </B><B> |     </B><A HREF="index.htm" TARGET="_top" TITLE="NETWORKS OF POSSIBLE FUTURES"><B>home</B></A></div> <div> <A HREF="index.htm" TARGET="_top" TITLE="NETWORKS OF POSSIBLE FUTURES"><B>NETWORKS OF POSSIBLE FUTURES</B></A><B>   |   </B><B>NETWORKS OF PREDICTIONS</B><B>   |   </B><A HREF="pivotal_predictions.htm" TARGET="_top" TITLE="PIVOTAL PREDICTIONS"><B>PIVOTAL PREDICTIONS</B></A><B>   |   </B><A HREF="five_themes.htm" TARGET="_top" TITLE="FIVE THEMES"><B>FIVE THEMES</B></A><B>   |   </B><A HREF="one_predictions_reach.htm" TARGET="_top" TITLE="ONE PREDICTIONS REACH"><B>ONE PREDICTION'S REACH</B></A><B>   |   </B><A HREF="a_similar_scenario.htm" TARGET="_top" TITLE="A SIMILAR SCENARIO"><B>A SIMILAR SCENARIO</B></A><B>   |   </B><A HREF="one_predictions_failure.htm" TARGET="_top" TITLE="CONSEQUENCES OF A FAILED PREDICTION"><B>CONSEQUENCES OF A FAILED PREDICTION</B></A><B>   |   </B><A HREF="an_alternative_future.htm" TARGET="_top" TITLE="AN ALTERNATIVE FUTURE"><B>AN ALTERNATIVE FUTURE</B></A><B>   |   </B><A HREF="betting_on_robustness.htm" TARGET="_top" TITLE="BETTING ON CONTINUITY"><B>BETTING ON CONTINUITY</B></A><B>   |   </B><A HREF="footnotes_and_updates.htm" TARGET="_top" TITLE="FOOTNOTES AND UPDATES"><B>FOOTNOTES AND UPDATES</B></A><B>   |   </B><A HREF="references.htm" TARGET="_top" TITLE="REFERENCES"><B>REFERENCES</B></A><B>   |   </B><A HREF="postscript_on_the_water_issu.htm" TARGET="_top" TITLE="POSTSCRIPT ON THE WATER ISSUE"><B>POSTSCRIPT ON THE WATER ISSUE</B></A><B>   |   </B><A HREF="id15.htm" TARGET="_top" TITLE="Florida Futures Survey"><B>Florida Futures Survey</B></A></div> <div> </div> </td> </tr><tr> <td valign="top" width=250 BGCOLOR="#336666" TEXT="#080000" background="063fab83.jpg"> <div>                             <IMG SRC="07676730.gif" border=0 width="118" height="115" ALIGN="BOTTOM" HSPACE="0" VSPACE="0">  </div> <div> <A HREF="index.htm" TARGET="_top" TITLE="NETWORKS OF POSSIBLE FUTURES">NETWORKS OF POSSIBLE FUTURES</A></div> <div> <B>NETWORKS OF PREDICTIONS</B></div> <div> <A HREF="pivotal_predictions.htm" TARGET="_top" TITLE="PIVOTAL PREDICTIONS">PIVOTAL PREDICTIONS</A></div> <div> <A HREF="five_themes.htm" TARGET="_top" TITLE="FIVE THEMES">FIVE THEMES</A></div> <div> <A HREF="one_predictions_reach.htm" TARGET="_top" TITLE="ONE PREDICTIONS REACH">ONE PREDICTION'S REACH</A></div> <div> <A HREF="a_similar_scenario.htm" TARGET="_top" TITLE="A SIMILAR SCENARIO">A SIMILAR SCENARIO</A></div> <div> <A HREF="one_predictions_failure.htm" TARGET="_top" TITLE="CONSEQUENCES OF A FAILED PREDICTION">CONSEQUENCES OF A FAILED PREDICTION</A></div> <div> <A HREF="an_alternative_future.htm" TARGET="_top" TITLE="AN ALTERNATIVE FUTURE">AN ALTERNATIVE FUTURE</A></div> <div> <A HREF="betting_on_robustness.htm" TARGET="_top" TITLE="BETTING ON CONTINUITY">BETTING ON CONTINUITY</A></div> <div> <A HREF="footnotes_and_updates.htm" TARGET="_top" TITLE="FOOTNOTES AND UPDATES">FOOTNOTES AND UPDATES</A></div> <div> <A HREF="references.htm" TARGET="_top" TITLE="REFERENCES">REFERENCES</A></div> <div> <A HREF="postscript_on_the_water_issu.htm" TARGET="_top" TITLE="POSTSCRIPT ON THE WATER ISSUE">POSTSCRIPT ON THE WATER ISSUE</A></div> <div> <A HREF="id15.htm" TARGET="_top" TITLE="Florida Futures Survey">Florida Futures Survey</A></div> </td> <td valign="top" height=390 width="100%" BGCOLOR="#FFFFFF" TEXT="#080000" background="06107073.jpg"> <div> <B>NETWORKS OF PREDICTIONS</B></div> <bR> <div> Most of the predictions in <I>Florida's Megatrends</I> were related to other predictions, creating chains of causal contingencies. </div> <bR> <div> To advance their arguments, Colburn and deHaven-Smith put forward a number of predictions - about a hundred in all. Some were explicitly stated. Others could be inferred from what they wrote. Most were related to other predictions, either directly or indirectly, creating chains of “if…then” causal contingencies. </div> <bR> <div> Some chains were linked together because certain predictions belonged to more than one chain. This interconnectedness created a network structure that was represented graphically, following<A HREF="http://en.wikipedia.org/wiki/Concept_mapping" TARGET="_top" TITLE="http://en.wikipedia.org/wiki/Concept_map"> the conventions of cognitive or causal mapping</A>.  (Novak and Cañas, 2006) </div> <bR> <div> The causal link between any two predictions was depicted by two nodes joined by an arrow indicating the direction of causal influence. A node with both incoming and outgoing links corresponds to a prediction that was both a cause and a consequence of other predictions.  </div> <bR> <div> Cognitive maps are directed labeled graphs and were used, in an influential study 30 years ago, to model the network-like structure of decision makers' causal beliefs about major policy issues. (Axelrod, 1976) Since then, this causal mapping technique has been used to format the content of knowledgebases, to diagram complex arguments (since cognitive maps are a type of argumentation), and as a tool for ramification analysis - the systematic tracing of potential outcomes of possible events. (Geffner, 1996; Hunter, 2000; Miao and Liu, 2000)</div> <bR> <div> The small network below shows the convergence of some of the economic and demographic changes that will spur the transition of Florida from a southern to a Sunbelt state.</div> <bR> <div> <IMG SRC="082dc310.jpg" border=0 width="988" height="561" ALIGN="BOTTOM" HSPACE="0" VSPACE="0"></div> <div> <FONT SIZE="1" COLOR="#333333" FACE="Arial" ><B>Comment:</B> Florida's economy and politics will more closely resemble those of other Sunbelt states (and less like those in the Deep South) if, as predicted, the growth of black and Hispanic populations will lessen the influence of whites, if Florida continues to attract large numbers of residents from elsewhere in the U.S., and if the expansion of international trade and the technology sectors reduces the state's dependence on tourism, agriculture, and construction. </FONT>(The network diagrams were constructed using Cmap Tools Knowledge Modeling Kit, version 4.10, developed at the <A HREF="http://cmap.ihmc.us/" TARGET="_top" TITLE="http://cmap.ihmc.us/">Institute for Human and Machine Cognition</A>, Pensacola, Florida.) </div> <bR> <bR> <bR> <div> <FONT SIZE="3" COLOR="#333333" FACE="Arial" >The distribution of links was markedly biased:  there were many nodes with few links and a few nodes with many. This bias is a well-known characteristic of non-random networks, regardless of their size, and is the reason why highly interconnected nodes give the network its structure. They hold it together. </FONT>(<FONT SIZE="3" COLOR="#000000" FACE="Arial" >Caldeira, et al., 2006</FONT>)</div> <bR> <DIV ALIGN="LEFT"></DIV> <div> MANY PREDICTIONS HAD A FEW LINKS AND A FEW HAD MANY</div> <bR> <div> <TABLE BORDER="2" CELLPADDING="1" CELLSPACING="1" WIDTH="601" BORDERCOLORLIGHT="#C0C0C0" BORDERCOLORDARK="#808080" FRAME="BOX" RULES="ALL" HSPACE="0" VSPACE="0" > <tR> <tD VALIGN=TOP COLSPAN=2 HEIGHT= 20 WIDTH="589"><div> <B>FREQUENCY OF LINKS BETWEEN PREDICTIONS</B></div> </tD> </tR> <tR> <tD VALIGN=TOP HEIGHT= 19 ><NOBR><div> <B>Number of direct links for each prediction</B></div> </NOBR></tD> <tD VALIGN=TOP WIDTH="293"><div> <B>Percentage of predictions</B></div> </tD> </tR> <tR> <tD VALIGN=TOP HEIGHT= 19 WIDTH="293"><div> 2 links </div> </tD> <tD VALIGN=TOP WIDTH="293"><div> 59%</div> </tD> </tR> <tR> <tD VALIGN=TOP HEIGHT= 19 WIDTH="293"><div> 3 links </div> </tD> <tD VALIGN=TOP WIDTH="293"><div> 30%</div> </tD> </tR> <tR> <tD VALIGN=TOP HEIGHT= 19 WIDTH="293"><div> 4 links </div> </tD> <tD VALIGN=TOP WIDTH="293"><div> 8%</div> </tD> </tR> <tR> <tD VALIGN=TOP HEIGHT= 19 WIDTH="293"><div> 5 links </div> </tD> <tD VALIGN=TOP WIDTH="293"><div> 3%</div> </tD> </tR> </TABLE></div> <bR> <div> <B>Note:</B> The links may be either in-directed, out-directed, or both.  Nodes with a single link (27% of the total) were excluded because these are either entry or exit points in a network with arbitrarily restricted boundaries. The network could be expanded at these points, adding at least one more link to each.</div> <bR> </td> </tr></table></body>