BACKGROUND

Research-practice transfer involves bidirectional communication between researchers and practitioners, while the research-practice gap refers to breakdowns in this communication. Network interventions are often viewed as a promising way to bridge this gap. However, the focus has primarily been on interventions designed to facilitate researchers' ability to push out (i.e., disseminate) their research.

PURPOSE

In this study, we propose three simple network interventions designed to facilitate practitioners' ability to pull in (i.e., search for) useful research: (1) relying on multiple sources of information to avoid dead ends, (2) relying on well-connected sources of information to avoid echo chambers and (3) relying on sources of information outside their own community to reach brokers.

METHOD & RESULTS

We use a simulation to evaluate the potential utility of the proposed interventions for improving the success and efficiency of research-practice transfer. The simulation suggests that all three interventions improve both the success and efficiency of practitioners' search for researchers in their social networks. Specifically, when a practitioner searching for research relies on multiple sources of information, relies on a well-connected source of information, or relies on a source of information outside their own setting, they are more likely to find a researcher in their network and they do so more quickly.

CONTRIBUTIONS

Taken together, these results suggest that simple practitioner-focused network interventions can improve research-practice transfer by increasing the success and efficiency of practitioners' searches for research.

NOTE: A preprint of the full paper can be downloaded at: https://osf.io/7u5ne_v1

The COVID-19 infodemic has underscored the challenge of health misinformation, significantly influencing public health decisions. Despite efforts to counter it, understanding the social mechanisms driving its spread remains a key research gap. This study integrates survey data with agent-based modeling (ABM) to analyze how misinformation propagates within different social network structures and its implications for public health.

A survey conducted in Spain (January–March 2024) gathered responses from 2,200 individuals, assessing health beliefs using the COVID-19 Misinformation Scale (CMS12). Factor analysis identified four misinformation dimensions (Conspiracy, Hoaxes, Vaccines, Fertility), which were used in a k-means clustering to classify respondents into three groups: Informed (49%), Hesitant (30%), and Misinformed (21%). After characterizing these social profiles via logistic regression, an ABM was developed to simulate misinformation spread in four network structures (Regular, Random, Small-World, Scale-Free) using igraph. The model incorporated network size, learning rates, and resistance to opinion change parameters.

Results indicate that misinformation spreads most effectively in scale-free networks due to the presence of highly connected hubs, while small-world networks tend to confine misinformation within local clusters. Hesitant individuals were found to be the most susceptible to misinformation. Overall, our findings suggest that the combination of a highly interconnected network, low resistance to change, and greater learning capacity facilitates the misinformation spreading among hesitant individuals—those with less defined opinions compared to the more stable and ideologically driven positions of the informed and misinformed groups.

These insights highlight the importance of targeting hesitant individuals in health interventions to mitigate misinformation’s impact.

Networks impact how we acquire information and adopt behaviors. This idea has long-standing theoretical roots, but only recently have researchers had the necessary data, computational power, and statistical methods to examine how contagion processes play out over different network structures. Current scholars of diffusion in networks argue that some structures facilitate the spread of some “infection” more than others, depending on whether the nature of object being diffused is “simple” and spread through mere exposure or “complex” and spread through social influence.

While both models highlight the importance of social structures on person- and system-level outcomes, they do so by reducing the actors within them to cultural dupes who are infected by information or behavior given some critical threshold of exposure or quantity of already-infected relations. This may be an artifact of the contagion model’s origins in communicable disease transmission, in which passive exposure is a sufficiently comprehensive pre-condition for becoming infected. But bits of social information and novel behaviors are not diseases. Actors who adopt cultural innovations are not dupes but individuals with agency, and often strategically seek out ideas and practices they might ultimately adopt through their social networks. Deliberate search behaviors – constrained or enabled by their network structure – likely pattern an actor’s chances of adoption.

We argue that current network contagion frameworks fail to capture a core aspect of human behavior: individuals often actively seek out and adopt cultural objects that are strategically beneficial or otherwise interesting to them. While human agents are indeed made aware of and influenced to adopt cultural innovations through their immediate social ties as these existing models propose, they also tend to seek out, encounter, and adopt these objects via foraging behaviors that provide additional opportunities for active self-exposure and deliberate social learning.

In our paper, we elaborate a set of speculative models that integrate the contagion literature’s insights about the influence of social network structure with the information foraging literature’s findings that people deliberately modify their strategies or the structure of their environments in the pursuit of obtaining valuable information or acquiring useful cultural tools. We then run a set of computational experiments that compare hypothetical rates of adoption in Erdős-Rényi graphs and locally clustered Small World graphs – structures that prior contagion research identifies as less and more beneficial to the facilitation of diffusion, respectively – in conditions where active information foraging strategies (i.e., approximations of the ACT-IF model of foraging) are or are not simultaneously parameterized alongside our manipulations to network structure and passive influence dynamics.

Our preliminary results contradict conventional wisdom in the contagion literature: we find that less clustered Erdős-Rényi graphs often facilitate more rapid and complete adoption of complex cultural objects than clustered Small World graphs when agents are incentivized to forage for information beyond their first-order social ties (explore) and strategically resample information when they encounter it (exploit). Our next steps involve using our hybrid foraging-contagion model to predict the adoption of new open-source libraries by developers on GitHub.

Purpose: Evaluate the interactions between the context delivery, mechanisms relating to peer leader influence and reduction in adolescent smoking.

Data: A Stop Smoking in Schools Trial (ASSIST) involved student nominated, peer leaders communicate smoke-free messages to peers in their school-year. Data archiving has recently extended availability of data to 53 schools across England/Wales (n=10,387).

Analysis: Stochastic actor oriented models (SAOMs) are estimated using ASSIST, to delineate dyad level influence and intervention effects. Interaction reduced friendship networks are modelled separately to examine interdependencies between multiple types of peer relationships. Using a two-stage analysis on networks, SAOM estimates are nested per period / per social focus. Meta-analysis is repeated for subgroups; setting (Welsh valley/non-valley) and intervention arm (ASSIST/control). The relative contribution of mechanisms to the social patterning of smoking is further examined so that proper attribution can be made.

Implications: SAOMs serve an important tool to build depth to mid-range theory, i.e. for whom and under what circumstances the ASSIST programme works and can be transferred to.

Authoritarian regimes are notoriously opaque in their decision-making processes and create challenges for external policymakers to understand and predict future behavior. Because of this lack of clear information, policymakers heavily rely on country experts who track developments on a day-to-day basis. These experts will have a strong potential influence on the policies of foreign actors towards the autocratic regime.

Our data focuses on China. Understanding Chinese politics has become increasingly important for policymakers as China has become a leading cultural, economic, and political global power. Experts in Chinese politics therefore play an important role as a source of information, analysis, and guidance.

While experts may directly observe politics in China, they also inform and are informed by each other. This creates an opportunity for the diffusion of knowledge and information within a network. However, it also creates a risk that sharing and amplifying inaccurate information or biased assessments within the network results in sub-optimal policy outcomes.

We use a snowball sampling process to accumulate information on over 2,000 individual China experts globally. The data contains self-reported assessments of their own expertise on Chinese politics from almost 500 respondents, and occupational background information from public sources on most of the nominated individuals. We also use the nominations to construct a network through which information and influence could potentially flow within this community.

We scrape blog-style publications of experts on salient issues in Chinese politics from online publications and blogs and apply topic modelling to investigate the nature of the information flow through the network. A Stochastic Actor-Oriented Model (SAOM) is used to estimate the level and shape of information diffusion across the network of experts. The findings indicate the presence of the diffusion of writings among networks of experts on Chinese politics, signifying a considerable potential for the existence of echo chambers among these experts.