Fig 1.
Theoretical contributions from Kania et al. (2018).
A conceptual triangle is hypothesized which brings out educated guesses on how common certain kinds of changes are likely to occur in relocation exercises. The wider the base, the easier to bring about that change. We notice, therefore, how structural changes are guessed to be easier to bring about than mental changes. We notice, however, two points. First, in this conceptual map, there is no backing from observed data. Second, the cosmetics do not reveal insights: that is, the base or the altitude of the component shapes do not hold special significance. Both these points are taken up in FIg 2, a realized version of this conceptual triangle through statistical metrics.
Fig 2.
Legitimacy of theoretical model for “Systems Change” from Kania et al. (2018) by theme according to data from RE.
! RE!NSTITUTE™ and mathematical concepts from Bhaduri (2023). We interpret this as a realized version on the theoretical triangle shown in Fig 1. The shapes and the sizes of the components have been calculated using data from the 100-day challenges. Unlike Fig 1, the bases and the altitudes of the shapes here do convey statistical information. The bases represent how common changes in a certain category are (the wider the base, the more common the change) while the altitudes represent the inverse of the length of the confidence interval for this statistic (the taller the altitude, the more sure we are in the value of the base). This observed empirical triangle (Fig 2), therefore, confirms, in the process, the general shape of the theoretical triangle (Fig 1) in that the bases for “practices”, “policy”, and “resources”, collectively (that is, “structural”) cover the widest ground and the bases subsequently taper down on relational and mental changes.
Table 1.
Support and confidence interval lengths summarized across features.
Table 2.
Data collected by the RE!NSTITUTE™. Each row represents one deployment of the 100-Day Challenge™. A cross indicates changes in the corresponding aspect could be brought about in that instance of the experiment.
Fig 3.
Frequency of change by theme according to data from RE.
!NSTITUTE™, based on a theoretical model for “Systems Change” from Kania et al. (2018) using mathematical concepts from Bhaduri (2023). The heights of the towers describe the rate at which changes in the corresponding types were witnessed to occur. For instance, of all the 100-day challenges studied, we found about 55% of times changes in “relationship” were observed, either changes in “relationship” only, or in “relationship” along with those in other types.
Fig 4.
A typical map of likely connections that lead to changes in mental structures.
The more likely connections are shown in deeper red. We invite readers to explore other possibilities (for instance, those connections that lead to non-mental targets) on our interactive dashboard. One implements these rules in an input-output sense. If changes in the input details – shown by the blue arrows – are promised, one is given the rate at which changes in the output detail – shown by the red arrow – accompany. The higher this rate, that is, the more (correlationally) automatic the input-output connections are, the darker the shade of the red node representing the rule. For instance, rule 12 says if one designs a relocation study that generates changes in “relationship”, “practice”, “power”, changes in “mental” structures will be promised with a reasonably high chance (50% if one hovers on the dashboard).
Fig 5.
Epochs of stable relocation rates are shown on either side of the red vertical separators for (a) Northern Illinois and (b) Desert City.
Each vertical separator represents an estimated change-point, that is, a point in time over the course of the 100 days during which the rate of relocating homeless people improved or deteriorated substantially. For instance, had the change-point around day 80 (=400/5) not been there for Northern Illinois, the total number of people relocated could have been 85 instead of 80. Crowding of change points towards the end of the experiment is observed for nearly all regions.
Table 3.
Records of days (within the 100-Day Challenge™ period) around which substantial changes occurred in the rate at which people were relocated.
Fig 6.
Network joining regions with similar change points.
If two cities experience drastic shift in relocation numbers around similar times over the 100 days, there will be an edge joining them. Else, there will not be one.