Communication & Phone Telemetry

1 Communication & Phone Telemetry

This section describes the longitudinal multi-modal communication and address book datasets passively collected from participant smartphones across 158 continuous academic study weeks (2011–2014).

All communication events have been programmatically de-identified: study peers are linked using standard 5-digit egoids, while out-of-network contacts (family, high school friends, non-study contacts) are protected using irreversible salted cryptographic pseudonyms (ext_XXXXXXXX and em_XXXXXXXX).

1.1 Download Datasets

1.1.1 1. Longitudinal Dyadic Communication Panel (Calls & SMS)

Dyadic interaction panel (\(N = 383,895\) dyad-week observations across 158 weeks) tracking directional phone calls and text messages exchanged between participants and their social contacts.

Download Compressed Dyads (.csv.gz, 2.7 MB) Download Dyadic Panel (.csv, 11 MB) Download R Dataset (.rds, 1.8 MB) Download Stata Dataset (.dta, 12 MB)

1.1.2 2. Longitudinal Weekly Communication Ego Metrics

Ego-level weekly summary panel (\(N = 19,598\) ego-weeks) tracking communication degree, total incoming/outgoing SMS volume, and total incoming/outgoing voice call duration.

Download Ego Metrics (.csv, 860 KB) Download R Dataset (.rds, 270 KB) Download Stata Dataset (.dta, 1.0 MB)

1.1.3 3. Longitudinal Email Ego Networks

Longitudinal email interaction panel (\(N = 676,717\) dyad-weeks from over 1 million parsed emails) classified into peer, personal webmail, campus listservs, and automated commercial notifications.

Download Compressed Email Dyads (.csv.gz, 3.7 MB) Download Email Dyads (.csv, 33 MB) Download R Dataset (.rds, 3.3 MB) Download Stata Dataset (.dta, 57 MB)

Download Ego Email Metrics (.csv, 520 KB) Download R Dataset (.rds, 140 KB) Download Stata Dataset (.dta, 680 KB)

1.1.4 4. Phone Address Book Rosters & Dynamics

Dyadic address book roster (\(N = 30,802\) saved contacts) and weekly ego metrics (\(N = 6,384\) ego-weeks) tracking contact retention and new social ties added to student smartphones.

Download Compressed Contacts Roster (.csv.gz, 210 KB) Download Contacts Roster (.csv, 820 KB) Download R Dataset (.rds, 150 KB) Download Stata Dataset (.dta, 920 KB)

Download Weekly Contacts Metrics (.csv, 340 KB) Download R Dataset (.rds, 75 KB) Download Stata Dataset (.dta, 460 KB)

1.2 Data Schema & Variables

1.2.1 Dyadic Communication Panel

Variable Type Description
egoid Integer Standard 5-digit student participant identifier
alter_id Character Alter identifier: 5-digit egoid if enrolled peer; salted pseudonym ext_XXXXXXXX if outside contact
is_study_peer Binary 1 = Alter is an enrolled NetSense peer; 0 = External alter
study_week Integer Continuous study week index (Week 1 = Aug 22, 2011)
stata_week Integer Stata calendar week index
flow_in Numeric Total incoming communication events from alter to ego that week
flow_out Numeric Total outgoing communication events from ego to alter that week
total_flow Numeric Total communication events exchanged (flow_in + flow_out)
is_saved_in_contacts Binary 1 = Alter was saved in ego’s smartphone address book; 0 = Unsaved contact

1.2.2 Weekly Communication Ego Metrics

Variable Type Description
egoid Integer Standard 5-digit student participant identifier
study_week Integer Continuous study week index
indegree Integer Count of unique alters who initiated communication to ego that week
outdegree Integer Count of unique alters ego initiated communication to that week
total_degree Integer Count of total distinct communication partners that week
sms_in Integer Count of incoming SMS text messages
sms_out Integer Count of outgoing SMS text messages
total_sms Integer Total SMS text messages exchanged
voice_in Numeric Total duration (seconds) of incoming phone calls
voice_out Numeric Total duration (seconds) of outgoing phone calls
total_voice Numeric Total duration (seconds) of phone calls
total_communication Numeric Multi-modal communication volume (total_sms + total_voice)

1.3 Sample Data Overview

Show Code 
library(readr)
library(dplyr)
library(knitr)

comm_sample <- read_csv("Data/weekly_communication_dyads_longitudinal.csv", n_max = 6, show_col_types = FALSE)
kable(comm_sample, caption = "Sample Rows: Longitudinal Communication Dyads")
ego_sample <- read_csv("Data/weekly_communication_ego_metrics_longitudinal.csv", n_max = 6, show_col_types = FALSE)
kable(ego_sample, caption = "Sample Rows: Weekly Communication Ego Metrics")
Table 1: Sample Rows: Weekly Communication Ego Metrics
egoid alter_id is_study_peer study_week stata_week flow_in flow_out total_flow is_saved_in_contacts
62560 176422 0 -1 2683 34 0 34 0
10060 143943 0 0 2684 0 1 1 0
10353 110922 0 0 2684 1 0 1 0
10353 111498 0 0 2684 1 0 1 0
10353 114644 0 0 2684 2 1 3 0
10353 122333 0 0 2684 2 2 4 0
egoid study_week stata_week indegree outdegree total_degree sms_in sms_out total_sms voice_in voice_out total_voice total_communication
62560 -1 2683 1 0 1 34 0 34 0 0 0 34
10060 0 2684 0 1 1 0 1 1 0 0 0 1
10353 0 2684 33 18 51 68 41 109 2 0 2 111
10841 0 2684 5 7 12 28 34 62 0 9 9 71
11692 0 2684 1 0 1 1 0 1 0 0 0 1
13116 0 2684 26 35 61 225 266 491 48 108 156 647