Source-linked AI summary
Time as a limited resource: Communication Strategy in Mobile Phone Networks
Giovanna Miritello, Esteban Moro, Rubén Lara, Rocío Martínez-López, Sam G. B. Roberts, Robin I. M. Dunbar
TL;DR
Because communication time is limited, the paper examines whether larger mobile networks constrain tie strength and whether heavier mobile use changes uneven time allocation. Using detailed records from 20 million users and 9 billion calls, it finds that large networks have weaker ties without proportionally more communication, while users consistently concentrate time on a few contacts.
Problem
Prior evidence on communication-time constraints and uneven time allocation relied mainly on questionnaire data, leaving patterns in large-scale actual mobile communication to examine.
Method
The study analyzes detailed mobile call records to measure network size, aggregated communication time, tie strength, and the distribution of time across contacts.
Results
Large networks do not receive proportionally more communication time and have weaker ties on average, while users of all mobile-use intensities allocate most time unevenly to a small number of contacts.
Takeaways & Limitations
The findings support time constraints on tie strength in large personal networks and a no-change model for how mobile-use intensity affects disparity.
Takeaways & Limitations
The study uses calls from a single mobile operator and does not establish whether mobile-use intensity causes network size or vice versa.
Abstract
from arXiv · showhide
We used a large database of 9 billion calls from 20 million mobile users to examine the relationships between aggregated time spent on the phone, personal network size, tie strength and the way in which users distributed their limited time across their network (disparity). Compared to those with smaller networks, those with large networks did not devote proportionally more time to communication and had on average weaker ties (as measured by time spent communicating). Further, there were not substantially different levels of disparity between individuals, in that mobile users tend to distribute their time very unevenly across their network, with a large proportion of calls going to a small number of individuals. Together, these results suggest that there are time constraints which limit tie strength in large personal networks, and that even high levels of mobile communication do not fundamentally alter the disparity of time allocation across networks.
I. INTRODUCTION
This study asks how mobile communication intensity and personal-network size shape the allocation of limited communication time across social ties. Using large-scale mobile records and null-model comparisons, it tests whether mobile use broadens networks, strengthens strong ties, or leaves uneven allocation unchanged.
- Motivation: Communication time constrains how many ties people can maintain at each level of emotional intensity.Mobile phones make communication easier and total mobile voice-call volume increased markedly in the UK between 2004 and 2009.
- Research question: The study examines whether mobile communication changes how people distribute limited time across their personal networks.It focuses on communication intensity, network size, and disparity in time allocation.
- Predictions: The paper considers whether greater mobile use broadens networks, deepens strong ties, or leaves the uneven distribution of effort unchanged.The three possibilities correspond to more homogeneous allocation, more uneven allocation, and no association between intensity and disparity.
- Approach: The analysis uses 9 billion calls from 20 million mobile users to relate communication strategies to network size and mobile-use intensity.The study treats differential time investment across connections as reflecting, to some extent, the value placed on relationships.
- Research gap: Previous mobile-network research largely examined whole-network processes or dyadic ties, while personal networks based on mobile records remained relatively understudied.Existing studies often covered only three best friends or a sample of 94 participants.
II. DATASET
The dataset is an anonymized mobile Call Detail Record network from a single European operator, covering more than 20 million users and 9 billion calls over 11 months.
- Data source: The study analyzes mobile Call Detail Records from a single European operator over 11 months.Records were anonymized, with users represented only by identification numbers rather than names or phone numbers.
- Scale: The database contains over 20 million users and 9 billion calls between users.Each call record contained several fields, although the passage does not enumerate them fully.
III. THE BOUNDARIES OF HUMAN COMMUNICATION
The study models mobile communication as a weighted social network and tests how total communication time and tie weights vary with network connectivity. Real networks show saturation and declining average tie strength at higher connectivity, unlike randomized networks.
- Network measures: A tie weight w_ij is the aggregated time users i and j spend talking, while connectivity k_i counts a user’s connected neighbors.Node strength s_i sums the weights of a user’s connections.
- Network structure: Mean connectivity is 85.2, with a median of 62 and a maximum around 500 connections.Connectivity and node strength both have highly skewed distributions.
- Strength versus connectivity: Real-node strength increases with connectivity and then saturates, indicating that highly connected users do not increase total communication time proportionally.The randomized network follows the uncorrelated approximation s_i = ⟨w⟩k_i.
- Tie-weight distribution: Tie-weight distributions show no significant dependence on social connectivity across connectivity quartiles.The figure compares four connectivity groups with the whole-population distribution and an exponential distribution sharing the same mean.
- Interpretation: For large connectivity, node strength grows sub-linearly, so users with many contacts spend less time than expected under random weight assignment.The mobile-network pattern differs from the super-linear strength growth observed in air-transportation networks.
- Tie strength: Average tie weight increases with connectivity initially, peaks around 20 connections, and then decreases as connectivity grows further.The peak remains stable across observation windows of 3, 7, and 11 months.
IV. TIME ALLOCATION DIVERSITY
The study finds that people with larger social networks spend less time on each connection, yet distribute communication unevenly across contacts. Disparity varies with network connectivity, but its scaling is largely reproduced by randomizing tie weights, implicating weight heterogeneity rather than communication organization alone.
- Network size and tie allocation: People with larger networks dedicate less time to each social connection, while tie-weight distributions show no appreciable dependence on social connectivity.Users typically allocate a small amount of time to many contacts and a large amount to a small number, regardless of social-circle size.
- Disparity measure: Disparity measures local heterogeneity, ranging from approximately 1/k_i under equal communication to 1 when one tie carries the node’s entire strength.The measure is sensitive to how unevenly tie weights are distributed.
- Disparity scaling: For k_i > 20, the observed relationship is fitted by k_iY_i ≃ k_i^α with α = 0.5, intermediate between perfect homogeneity and perfect heterogeneity.Because α is below 1, disparity depends on social connectivity.
- Randomization test: Randomizing tie weights reproduces the same disparity scaling, although real communication is slightly more homogeneous than the randomized version at a given connectivity.This suggests that the scaling is primarily a reflection of the long-tailed distribution of tie weights.
V. CONCLUSIONS AND DISCUSSION
Across mobile users, larger personal networks do not receive proportionally more communication time, and tie strength declines beyond roughly 40 connections. Regardless of mobile-use intensity, users allocate time unevenly, concentrating communication on a small number of strong ties.
- Disparity: Real-network disparity is slightly smaller than after randomizing tie weights, but the scaling of disparity shows no significant difference between the two.The figure places observed data between perfect homogeneity and perfect heterogeneity, with an exponent of 0.5 for large connectivities.
- Network size and time: After approximately 40 connections, average tie strength decreases as network size grows.For smaller networks, time devoted to each connection grows proportionally with network size.
- Disparity: Users with high and low mobile-use intensity both distribute time very unevenly, devoting much of it to a small number of contacts.Mobile communication does not appear to fundamentally change this allocation pattern or strategically build weak ties.
- Study scope: The study analyzes 20 million users and 9 billion calls over almost one year, providing communication records at a scale unavailable to questionnaire-based studies.Its coverage is described as highly generalizable to the European country where the mobile company operates.
- Study scope: Because only calls between customers of one operator were included, measured personal networks may not represent users’ complete personal networks.The analysis therefore omits contacts using other mobile providers.
- Study scope: The study does not determine whether greater mobile-use intensity produces larger networks or whether pre-existing larger networks produce more phone use.The direction of this causal relationship remains unresolved.