Evaluate whether there is evidence for a strategy’s assumptions.

Is there a quan­ti­ta­tive or qual­i­ta­tive assess­ment that mea­sures the spe­cif­ic assump­tion in the link? If yes, then there is evi­dence for this link. If no, there is not evi­dence for this link. The pur­pose of this step is to iden­ti­fy whether evi­dence exists for the actu­al assump­tion, not its poten­tial. Evi­dence can come from quan­ti­ta­tive meta-analy­sis, key infor­mant inter­views, tra­di­tion­al knowl­edge, or many oth­er sources here as in oth­er uses of evi­dence, but it must be used to assess the spe­cif­ic assumptions.

For exam­ple, con­tin­ue the wind results chain, and con­sid­er the first assump­tion that util­i­ty require­ments for com­pa­nies to avoid sen­si­tive grass­land areas lead to util­i­ties pur­chas­ing pow­er only from those com­pa­nies that do so. Evi­dence for this assump­tion may come from util­i­ty pow­er pur­chase records (quan­ti­ta­tive), or from inter­views with man­agers from wind pow­er pro­duc­ers (may be quan­ti­ta­tive or qual­i­ta­tive). If in man­ag­er inter­views, man­agers say they think util­i­ties will pur­chase more wind pow­er from them in the future after they change prac­tices because they have a good rela­tion­ship with the util­i­ties, this does not con­sti­tute evi­dence. If, how­ev­er, man­agers say they have had high­er sales to the util­i­ties after improv­ing sit­ing prac­tices, that qual­i­ta­tive infor­ma­tion con­sti­tutes evi­dence for the assump­tion because it reflects a direct obser­va­tion that the assump­tion is true.

In the process of eval­u­at­ing evi­dence, teams will like­ly find they need to hone the lan­guage in the results chain, to be more spe­cif­ic about assump­tions and causal mech­a­nisms of change that are implic­it in the dia­gram. For exam­ple, con­sid­er the fol­low­ing exam­ple from north­ern Australia:

Often, results chains are word­ed ambigu­ous­ly. Thus, a first step in assess­ing the evi­dence for a results chain is to be re-word the results chain so that it is clear what evi­dence is required. The revised results chain below is more specific.

Note that even a spe­cif­ic results chain will require def­i­n­i­tion of terms. In this exam­ple, it must be clear to read­ers what is meant by ‘ear­ly sea­son burn­ing’, ‘veg­e­ta­tion com­mu­ni­ties of con­ser­va­tion con­cern’ and ‘native fau­na’. Evi­dence search­ing will be eas­i­est when results chains are high­ly specified.