The 30-minute caller text contrast between missed calls and calls with positive duration was uncertain in four weeks of Copenhagen student records

We tested caller text activity before and after calls in four weeks of public Copenhagen student records. The main window was 30 minutes. A call entered the sample only if no call on the same pair started in the previous 30 minutes. The two-status sample had 230 missed calls and 1,670 calls with positive duration. Zero-duration calls were kept separate. A within-pair comparison used 149 missed and 375 positive-duration calls on 93 ordered pairs. The missed-minus-positive difference in the after-minus-before text rate was 7.19 percentage points. The descriptive 95% interval was -1.44 to 15.66 points with pair sampling. It was -6.81 to 21.68 points with shared person weights. A dyadic-cluster variance check gave a similar interval to pair sampling. These data do not resolve the contrast and do not establish that it is zero. Fixed point estimates were positive under other windows and rules, but estimates changed sign between the first and second halves of the record. Those subsets can use different pairs. The study does not identify causal replies or message intent. Separate event counts, a full indicator regression, and an explicit variance check agreed. A complete repeat reproduced seven numeric files exactly. The package preserves the source data, provenance, local plan, code, random draws, figures, and reproduction steps.

human communication · missed calls · SMS · temporal networks · cross-channel communication · uncertainty · Copenhagen Networks Study · reproducibility