Missed calls precede more one-hour callbacks in four weeks of Copenhagen student call records

We tested whether missed calls precede more rapid calls in the reverse direction. We used 3,600 public call records from four weeks of the Copenhagen Networks Study. The file has 536 user IDs. A callback was the next call on the same pair, in the reverse direction, within one hour. Calls marked as missed had 121 callbacks among 366 eligible calls, or 33.1%. Calls with positive duration had 389 among 2,650, or 14.7%. The raw rate ratio was 2.25. The difference within the same caller–recipient pairs was 22.1 percentage points. It used 145 ordered pairs with both call statuses. The descriptive 95% interval was 15.6 to 29.0 points with pair sampling. It was 11.8 to 33.7 points with person weights. The difference remained with added time controls, with other windows, and after callbacks in the first two minutes were excluded. Zero-duration calls were kept separate. Source hashes, separate count and model checks, and a full repeat passed. The package preserves data, code, provenance, figures, results, and repeat instructions. This exploratory result concerns recorded calls in one student network. It does not show that missed calls cause callbacks or measure trust or cooperation.

temporal networks · phone calls · reciprocity · missed calls · Copenhagen Networks Study · open data · reproducible research