A correction supplies evidence about a claim. The replies and quotations it prompts can also bring the claim back into circulation. Both effects belong to the same exchange. Relational intervention changes the timing, access and connections through which further exchanges become possible. Disinformation is one part of this wider field of information.
Identify the Dependency
Start with an observable relation. A response can renew attention to an earlier item, and that renewed attention can generate another round of replies. Establish which links carry the return, the intervals between encounters and the conditions in which the sequence recurs. A shared news event or a regular schedule can also bring several streams into step. The comparison begins with measurement and diagnostics.
A change in timing alters when a response becomes available relative to other events. A change in access or recommendation alters the routes through which it is encountered. These changes can affect the same dependency. Moving an item in a sequence changes both when it is seen and what is likely to precede it. Specify which part of that relation the intervention acts on and what consequence would support the explanation.
Test a Change in Timing
In the proposed frequency domain experiment, change the time at which a response is released while keeping its wording fixed. Choose the exchanges in advance and specify each response, its source and its delivery path before assignment. In a trial, assign reference and shifted schedules by chance to whole exchanges. Record actual release and reception where available. The assigned delay and the interval a participant experiences are different observations.
Choose a common follow-up duration h from each assigned release. A possible outcome is a new reply to the earlier item followed by a reply to that reply, both within the window. For each schedule g, let Mg count the eligible items assigned to it and Ng(h) count the items for which the sequence occurs at least once. With complete follow-up and Mg > 0, the ratio gives the observed recurrence fraction.
rg(h) = Ng(h) / Mg
Compare the fractions under the two schedules and retain the assigned groups when reporting the result. Measure exposure separately, since a change in timing can change who encounters the response. Record gaps in observation. Compare exchanges facing similar external events and track links that carry activity between the groups. These crossings are part of the effect to be examined.
A companion comparison varies wording at the same assigned release time. Timing can change the significance of identical wording, so examine how each response is taken up. The delayed copy calculation provides a reference for phase displacement. In the exchange itself, participants can respond differently and create new paths. Measure the resulting timing, links and recurrence.
Follow the Effects Through the Field
Changing a route of encounter can change what is available for another round of communication. An item can remain accessible while a different sequence of recommendations brings other material into view first. Observe whether replies return by other routes, whether their timing changes and whether the relation under study continues across the change.
Evaluate the outcome against the purpose stated before the comparison. If the aim is to interrupt a recurring reply sequence while keeping a correction available, record both the sequence and access to the correction. Follow the effect over the relevant intervals. A decline in one channel can accompany renewed activity elsewhere. Delays, verification work and demands on participants can also move through the network. Include those effects in the assessment.
Institutional Feedback and Provisional Control
Control is always provisional. A rule that changes visibility, sequencing or access becomes part of the conditions under which people communicate. Their responses change the dependencies on which the rule acts. Institutions participate through their classifications, procedures and decisions about which consequences count as evidence. See communication and regulation.
Review has to reach the rule itself. Record the basis for an intervention, the effects expected and the observations that warrant revising it. A procedure that protects its own measure of success can lose sight of changes elsewhere. Evidence from affected channels and participants gives the institution a way to correct the conditions it helps produce.
The chapter on operational implementation takes up how to carry those observations into models, decisions and subsequent review.
