Selecting Materials Outreach Topics from Community Repair Demand: Organizational Weighting, Temporal Transfer and Local Fit

Community repair events bring materials science into contact with familiar products, but outreach teams must decide which objects deserve limited preparation time. This study evaluates topic selection under two representation rules: equal influence for each recorded repair attempt and equal influence for each repair group. The Open Repair Alliance collection contains 48,669 attempts. Selection used 17,805 attempts from 95 groups, followed by evaluation on 18,802 attempts from 126 groups. The four topics selected by attempt weighting were laptops, lamps, separate audio equipment and coffee makers. Group weighting replaced the latter two with mobile phones and small kitchen appliances. Its selection-period advantage in average group coverage contracted from 3.09 to 0.43 percentage points in the following year. Network-stratified group resampling gave a 95% interval of −1.29 to 2.28 points for that contrast, while attempt coverage decreased by 3.27 points. Local differences had opposing signs: 58 groups gained coverage, 63 lost coverage and five were unchanged. Resampling the selection groups recovered the complete topic sets in only 19.28% and 35.84% of draws. Increasing the kit to eight topics made the two rules select identical categories; ten topics were needed to exceed one-half of average group demand. These results establish the consequences of representation choices without attributing learning gains to category matching. A shared materials kit therefore requires an explicit intended beneficiary, host-specific preparation and an educational assessment independent of repair completion.