Your task
Start without advance reading. Use the explanations inside the activity. You advise the Civic Technology Commission in fictional Bridgehaven. A proposed civic chatbot has passed one inspection, but the commission must decide which evidence and conditions make its use acceptable. Your job is to read two research studies, inspect dated records, audit exact statements and defend a rule within the institution’s powers and audit budget. You will also test your reasoning in another public-service setting. Keep a running distinction between what a document records, what you infer and what remains a proposed causal link. Your account and rival must predict different observations for the same outcome and population. A strong answer explains a finding supported by the evidence as well as a limit. Several political choices can earn full credit when feasible and well defended. Complete the record in the app, revise your earlier notes into a brief, and export your work for separate submission.
Explain the studies using the in-app explanations
Article A · Explain the study
Use the Article A explanation here; no journal reading is required.
1. Describe the four conditions: human or AI, each with or without demographic information.
2. Which comparison changes only access to a profile? Which also changes whether the opponent is human or AI? What agreement response was measured, among whom?
3. What does the correction change, and which positive comparison remains? Use relative odds as the unit, not a percentage of people persuaded. Exact statistics need not be memorized.
4. Name a limit and explain why uncertainty about persuasion does not settle every privacy question.
Article B · Explain the study
Use the Article B explanation here; no journal reading is required.
1. Choose one training or instruction comparison. What changed, among whom, and what response was measured?
2. Explain a finding and a limit. Several features of a message can change together, so a package effect does not identify which feature did the work.
3. How does the separate accuracy or fabrication finding help you avoid an unsupported explanation?
Use both studies
Use the two article explanations here; no journal reading is required.
1. Explain how the first study’s uncertain added-profile result and the second study’s small positive personalization result could both occur. Compare what changed, who took part, and what was measured.
2. What does using both studies change in your claim about personalization or your proposed audit?
3. Which numbers use different units and should not be subtracted? What direct comparison and uncertainty information would make your next claim stronger?
Before you start
No advance student reading is required. Use the simplified explanations inside the activity, with time for questions, analysis and saved continuation across sessions.
Open The Persuasion Hearing and choose Class for the required inquiry. Check your instructor’s assignment instructions; Fun exploration is a separate route.
Prepare a place for notes and downloaded backups. Keep the reading packet available so you can check a comparison, definition or source detail while writing.
For each study, note the two conditions, changed feature, population, measured outcome and unit. Bring one question about what its design leaves unresolved.
Work through the inquiry
1 · Open the hearing
Choose a focus for the opening puzzle and sketch your first account if useful. Identify an actor, an action and a possible outcome. Later, before opening raw evidence, you must commit your own explanation, rival and contrasting prediction. An initial guess earns no special credit for being right.
2 · Reconstruct both studies
Read and mark both explanations. Explain Article A’s four conditions and distinguish its two contrasts. For Article B, choose an assigned training or instruction comparison. Keep each population, outcome and unit attached to its comparison. Explain a concrete possibility that makes the studies compatible and changes an conclusion; two citations alone do not do this work.
3 · Inspect and revise
Commit your account, rival, prediction and first record choice. Open the required raw records, write a provisional finding, then open the later inspection. Mark observed links separately from hypotheses. Choose one link for the rival test and keep its outcome and population fixed. Retain, narrow, change or suspend your account with reasons; your earlier submission stays in the record.
4 · Audit, decide and interpret
Complete all six factual-audit rows with a status, reference or missing evidence, and reason. Defend the separate value judgment. Choose a licensing package or adaptation within the stated 120-hour budget; specify authority, pass evidence, enforcement, usable complaints and a legitimate burden. Read the exact plots and tables, submit the requested rationales before feedback, and explain your numerical conclusion.
5 · Apply the reasoning independently
Use the new institution’s own dossier. Complete the offer/completion ledger and name your target outcome, population and causal link. Keep nonusers in an offer comparison. Commit your account and predictions before the later record. Then defend or revise them and design one comparison with differing predictions. Preserve complete official information and usable multilingual complaints in every comparison group; name what your test leaves unresolved.
6 · Synthesize and submit
Deliberately carry your own source and decision notes into the brief, then revise them for the final argument. Explain an actual changed conclusion or justified retention, a positive finding that remains and a new uncertainty. Review units, references, costs, authority and the rival comparison. Complete self-review, print or export your record, then submit it separately through your course’s Canvas assignment.
Vocabulary
Component contrast
A comparison intended to change one feature while keeping relevant features comparable. Name the conditions before attributing an effect.
Package effect
An effect of an assigned bundle of features. The bundle’s effect does not identify each feature’s separate contribution.
Relative odds
A ratio of odds across conditions. Binary odds compare an event’s probability with its complement. Article A uses ordered agreement and an ordinal model; its headline is not a percentage of people persuaded.
Score points
Additive units on a specified score scale. A score-point difference is distinct from an odds ratio or a percentage of people.
Confidence interval
An interval supplied by an estimation procedure for a named contrast. Read its endpoints and unit alongside the point estimate.
Step connecting the cause to the result (mediator)
A proposed step through which a cause affects an outcome. This differs from a person or system facilitating discussion.
All-assigned offer effect
The outcome contrast retaining everyone in the original offer and comparison groups, including people who never use the service.
Selection into completion
Finishers may differ from nonfinishers before using a service. Voluntary completion is distinct from random assignment to an offer.
Not established
The available record does not settle a factual claim. This differs from evidence contradicting it; a value judgment needs a defended standard.
Contestability
People can see a decision’s reasons, challenge it through a usable route and obtain empowered review. Language and accessibility affect whether that route can be used.
Read and explain the graph
Read where the evidence came from, labels, axis, outcome, unit, population and follow-up period. Write the base or denominator beside each value before calculating.
Write the two named conditions and the subtraction order. Ask whether the comparison changes a single component or a broader package.
For an interval, identify its point, endpoints and zero. Use the supplied direct contrast’s uncertainty; subtracting separate interval endpoints does not derive it.
Show calculations and weights with their units. Separate descriptive sample differences from estimates with supplied uncertainty, then state which additional observation your political conclusion needs.
Six assessment criteria
Your instructor reviews these six actual criteria once each on a 0 to 3 scale, for a maximum of 18. Completion and factual checks record work; they do not grade the argument. A supported retention of an earlier account can earn full credit.
1. Source contribution and bounded positive findings
2. Exact sentence verification
3. Process, same-target rival and core revision
4. Lawful feasible licensing and usable challenge
5. Numerical comparison, population and precision
6. Unfamiliar transfer and protected study design
The shared progression is 0 for missing or materially inaccurate work, 1 for emerging reasoning, 2 for developed reasoning, and 3 for convincing, specific and bounded reasoning. Your evidence, assumptions and feasible comparison matter; political agreement, prose polish, help use and changing your mind are not grading criteria.
Save and export your work
Check the app’s save status. Browser progress is kept on the current device when saving is available; keep using the same browser to resume that local run.
Download a progress JSON backup at the end of each work session and keep it with a clear activity/date label. This backup preserves editable work and submitted histories.
Use the app’s restore/import control with a backup from this activity. Earlier-version notes may appear as preserved history while revised tasks still require new submissions.
If the app reports another saved version, download your current draft before choosing which version to keep. Saving progress, or using a configured class connection, is separate from submitting the required Canvas files.
Optional private online saving
Online saving is optional. Connecting to a saving service does not upload or submit the device draft. Use Save online personal draft only when you want to create or update that personal copy; keep its private recovery key and your local JSON backup.
A personal online draft is not an instructor submission. To submit through a configured hosted class, find the class with its student submission code, review the destination, apply the assignment switches if requested, complete the Class record and save the current version online. Tick the explicit consent box and use Submit this response snapshot.
Submission sends that version to the identified class. Later edits stay personal until you explicitly submit again. This hosted-class route is separate from the Canvas assignment and its upload confirmation.
Submit separately to Canvas
Use Print / save as PDF or the brief text export to create the record your instructor requests. Check that it includes your source notes, audit rows, decision, numerical work, unfamiliar-case revision and final reflection. Upload that file to the separate Canvas assignment and complete Canvas’s submission action. Keep the progress JSON as a backup; submit it only if requested. The app’s completion flag and local save do not submit work to Canvas.