A Data Science Central Community
Leveraging grasstops (i.e. group leaders) is imperative given their highly influential position with other group members. If a grasstop can be prevented from terminating legislative (i.e. policy oriented) support, the legislative termination rate for the group members has a higher likelihood for reduction. Alternatively, attempting to get a grasstop from a competing policy issue to switch or terminate may increase likelihood of similar results with group members associated with that competing policy issue while reducing the legislative termination rate of group members associated with your legislative support agenda.
Interestingly, many approaches to analytics primarily focus on individualism, in fact, how many studies involving research relating to capitalism does the same? Basically, they use of a variety of data about individuals to generate key indicators of the behavior to make interpretations or predictions. If any individual has values for the key indicators that are associated with the occurrence of the behavior, that individual can be targeted as a person of interest based on behavior, or in advanced cases, attitudes.
Consider approaches to explore legislative support as outlined in the first paragraph, in which a citizen terminates a legislative relationship with an enterprise. Clearly, the cost of retaining this support is significantly lower than the cost of cultivating a new one; therefore, the ability to identify citizens at risk of termination is included in the bottom-line critical category. Commonly, some personnel (i.e. lobbyist, analyst, technologist, or policy maker) often rely on a number of key performance indicators (KPI) to describe these legislative supportive citizens (i.e. grassroots), including demographic information and recent legislative support interaction logs (i.e. calls, visit to the capital, complaints to legislator, etc.). Predictive Analytics initiatives based on these attributes use changes in legislative support patterns that are consistent with interaction patterns of citizens who have terminated support in the past to identify citizens having an increased flight risk. Ideally, citizens identified as being at risk receive additional lobbyist visits, public announcements, or other legislative support service options in an effort to reduce flight risk or assure support retention.
Ironically, these methods commonly overlook other key social information that may significantly affect the behavior and attitude of a citizen. For instance, information about community synergy, public problematic issues, or key interactions amongst sub-groups of community members. As a result, linkage to a grasstop relationship with others would grant insight and influence. This influence can affect a grassroots' decisions, behaviors, and attitudes. Furthermore, if analyses that include only individual measures are omitting these important factors then predictive capabilities are restricted.
Fortunately, some service providers have produced Business Intelligence & Analytics tools, applications or systems (TAS) to assist with this demand for grasstops analytics. For instance, IBM® SPSS® Modeler Social Network Analysis addresses this problem by including features for processing information based on multiple relationships into derived fields that can be included as part of ensemble (i.e. linkage of individual or group of models) models.
Alternatively, some authors provide some books that may help as well;
In sum, seek TAS with features and functions that allows the ability to measure social characteristics (e.g. group-based) for individuals and the linkage of these social properties with individual-based measures in such combinations that a better overview of individuals relationship are included. Consequently, the results can improve the predictive accuracy of your initiatives.