Next: A Test of the Up: Exposure, Networks, and Mobilization: Town Previous: A First, Visual Insight

Defining the Boundaries for the Statistical Analysis

For the statistical tests we need to define more precisely the boundaries of our population. Since we are dealing with partly (or potentially) incomplete historical sources that report behavior, we will restrict our analysis to the part of the population that is better documented. Out of the two sources of data, the petitions and the structural information, the petitions are complete historical sources whereas the co-participation data may be less complete.

Therefore we will define the boundaries of the population on the basis of the petitions and include additionally all those individuals for whom we know that they have not signed a petition but are structurally integrated. Assuming that exposure is related to protest, these non-petitioners should be far less exposed to activists.

From the 5097 inhabitants of Esslingen mentioned in various historical records, 2184 have signed at least one of the petitions. Out of these petitioners 1284 can be localized within the structure of the city events. For 900 persons who signed a petition this is not possible since they are not listed in combination with any of the events we know of. Hence, for these 900 petitioners we can not show how their embeddedness in the structure and their exposure drove them to become politically active. Since we do not know whether this lack of integration is a lack of knowledge or an empirical fact we will have to exclude them from the analysis. On the other hand 702 persons are described through the events but did not sign any of the petitions. Since we definitely know that they have not signed a petition they must be included into the analysis. Therefore, we can test the degree of exposition to activists cross-sectionally at each of the three points in time and identify the extent to which this exposure influenced them to become personally active for 1986 persons.