PFPedro Ferreira

Pedro Ferreira

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article

Previous Organizations / Institutes

The organizations / institutes from the scholar's other articles in the past ten years


Pedro Ferreira has a total of 31 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    Multimodal Measurement of Cognitive Load in Instructional Videos

    by Yi-Yun Chu, Pedro Ferreira, Michael D. Smith, Mi Zhou

    Digital learning platforms have become a fundamental interface between learners and educational content, making it critical to understand how platform design shapes learner engagement and performance. This study investigates how video and instructor characteristics, cognitive load, and learner behavior interact to influence engagement and academic outcomes in online learning environments. We develop a multimodal, theory-guided pipeline that measures cognitive load directly from instructional video content, extracting visual, auditory, and textual features and mapping them to intrinsic, extraneous, and germane load. Applying this pipeline to a unique large-scale dataset from Outlier.org, spanning over 1,200 instructional videos and more than 14,000 learners across 14 subjects, we examine how different components of cognitive demand relate to learner engagement and performance. Our results show that while learners exhibit demographic preferences in instructor selection, these preferences do not reliably translate into better performance; instead, content-based cognitive load, measured directly from video, is a robust indicator of engagement and learning, highlighting the potential for real-time, AI-driven interventions to provide adaptive scaffolding. Our findings offer actionable insights for platform designers and educators, advocating for dynamic, modular video content and behavior-aware personalization to enhance learning outcomes at scale.

  • Open Access

    The Dynamic Aspects of Cognitive Load in Educational Videos

    by Yi-Yun Chu, Pedro Ferreira, Michael D. Smith, Mathieu Bonzon, Matt Gibson

    In this paper, we study how intrinsic, extraneous, and germane cognitive load relate to learning outcomes throughout instructional videos. We conceptualize instructional videos as configurable digital instructional artifacts whose instructional dimensions are allocated throughout time and construct measures of realized instructional exposure by combining cognitive load representations with learner interaction histories. Using observational data from an existing educational platform together with a complementary randomized instructional experiment, we examine how instructional characteristics unfold throughout instruction and relate to educational outcomes across multiple levels of educational aggregation.Across observational analyses, aggregate relationships between cognitive load and learning outcomes broadly align with established predictions from Cognitive Load Theory. However, these relationships vary systematically throughout instruction. Instructional experiences characterized by higher intrinsic load earlier in instruction tend to exhibit more positive relationships with learning outcomes than comparable instructional experiences in which intrinsic load occurs later. Measured extraneous load exhibits comparatively more positive relationships with learning outcomes during introductory instructional stages and less positive relationships later in instruction. Measured germane load exhibits increasingly positive relationships with learning outcomes later in instructional experiences.To complement these observational findings and introduce experimentally assigned instructional variation, we conduct a randomized experiment comparing alternative instructional video implementations that preserve conceptual content and learning objectives while varying instructional sequencing and instructional support. Experimental results identify causal effects of instructional organization that vary across learners and exhibit temporal patterns broadly aligned with those observed in the observational setting. Additional instructional support appears particularly beneficial for learners with lower prior familiarity with chemistry content, whereas more condensed instructional implementations perform comparatively better for learners with greater prior familiarity with chemistry content.Together, these findings suggest that instructional effectiveness depends not only on which instructional characteristics are present but also on when they occur throughout instruction. More broadly, the consistency of these temporal patterns across observational and experimental settings suggests that temporal allocation represents an important, measurable, and potentially optimizable design dimension of instructional videos and provides a foundation for adaptive educational systems that personalize not only what learners receive but also how instructional experiences unfold over time.

  • Top Cited Papers

    From Smartphones to Smart Students: Learning vs. Distraction Using Smartphones in the Classroom

    by Zhe Deng, Zhi (Aaron) Cheng, Pedro Ferreira, Paul A. Pavlou

    This study evaluates the impact of classroom smartphone use on student performance through two large-scale randomized controlled trials in China. Students were randomly assigned to one of four experimental conditions: (i) smartphones banned, (ii) smartphones allowed and used at will without guidance, (iii) smartphones allowed with teacher prompts to use them for instruction, and (iv) smartphones banned with teachers prompting the use of paper-based aids. Our findings show that unstructured smartphone use reduced performance compared with banning them, whereas teacher-guided use significantly enhanced learning outcomes. Paper-based aids yielded no measurable performance gain over the ban. We analyzed classroom video recordings to track individual-level time spent in each condition. We found that students spent similar total time learning across conditions, but teacher-directed smartphone use produced disproportionately large marginal learning gains outweighing the losses from distraction. Guided use also helped close gender and performance gaps though it risked widening digital divides by major and region. These results suggest educators should receive support to develop purposeful, app-based smartphone instruction materials; policymakers should avoid blanket bans and instead adopt structured usage policies; and tech developers should design classroom-friendly tools aligned with instructional goals. With thoughtful implementation, smartphones can become engines of equity, positive engagement, and learning within the classroom.


  • When Less Is More: Content Strategies for Subscription Video on Demand

    by Miguel Godinho de Matos, Samir Mamadehussene, Pedro Ferreira

    A field experiment shows that gradual content release boosts engagement and retention more than binge release on a streaming platform.


  • (How) Does User-Generated Content Impact Content Generated by Professionals? Evidence from Local News

    by Ananya Sen, Tom Grad, Pedro Ferreira, Jörg Claussen

    Many platforms host user-generated content (UGC) and content developed by professionals side by side. However, thus far, their impact on platform ecosystems has been mostly studied in isolation. In this paper, we use data from a network of 122 local news outlets hosted by an online news platform to study the spillover effects from UGC developed by citizen journalists to the content developed by professional journalists. We use the removal of a status index associated with citizen journalists as an exogenous shock to their supply of UGC to identify these spillover effects. We find that experienced citizen journalists reduce their production of content when this status index is removed. We then find that inexperienced professional journalists increase their output in response to this behavior. However, as a result of these changes, we find a reduction in the overall content hosted by the platform, especially in the case of local news and in more isolated regions. We further show that this is likely to have detrimental effects for the platform. In particular, there is a decline in overall viewership, and the platform may need to hire and pay salaries to more professional journalists to produce enough articles to close the gap left by the departing citizen journalists. Our work contributes to the literature on UGC and online platforms and to the literature on local news. This paper was accepted by D. J. Wu, information systems division. Supplemental Material: The online appendix and data files are available at https://doi.org/10.1287/mnsc.2023.4962 .


  • Open Access

    When Less is More: Content Strategies for Subscription Video on Demand

    by Miguel Godinho de Matos, Samir Mamadehussene, Pedro Ferreira

    In this paper, we devise an analytical consumer choice model in which consumers can visit a subscription video-on-demand platform to watch content from their watchlist and browse the catalog to search for additional content. We also partner with a large multinational telecommunications provider and implement a randomized field experiment that manipulates the timing and quantity of how much content is available for consumers to watch. Our model and experiment rationalize why content distributors are returning to drip-style content release schedules that limit how many content items consumers can watch. This apparent industry pullback contrasts with the popularized Netflix strategy of all-at-once from the beginning content release approach. We show evidence that drip-release content strategies induce more search moments, increasing the probability of matching content in the platform's library and consumer preferences. Ultimately these additional search moments trigger more utilization of the subscription video-on-demand system and continued subscriptions over time.

  • Open Access

    Consumer Behavior in the Online Classroom: Using Video Analytics and Machine Learning to Understand the Consumption of Video Courseware

    by Mi Zhou, George Chen, Pedro Ferreira, Michael D. Smith

    Video is one of the fastest growing online services offered to consumers. The rapid growth of online video consumption brings new opportunities for marketing executives and researchers to analyze consumer behavior. However, video introduces new challenges. Specifically, analyzing unstructured video data presents formidable methodological challenges that limit the current use of multimedia data to generate marketing insights. To address this challenge, the authors propose a novel video feature framework based on machine learning and computer vision techniques, which helps marketers predict and understand the consumption of online video from a content-based perspective. The authors apply this frame- work to two unique datasets: one provided by Masterclass.com, consisting of 771 online videos and more than 2.6 million viewing records from 225,580 consumers, and another from Crash Course, consisting of 1,127 videos focusing on more traditional education disciplines. The analyses show that the framework proposed in this paper can be used to accurately predict both individual-level consumer behavior and aggregate video popularity in these two very different contexts. The authors discuss how their findings and methods can be used to advance management and marketing research with unstructured video data in other contexts such as video marketing and entertainment analytics.


  • The Effect of Binge-Watching on the Subscription of Video on Demand: Results from Randomized Experiments

    by Miguel Godinho de Matos, Pedro Ferreira

    We analyze outcomes of two randomized field experiments to study the effect of binge-watching on subscription to video on demand. In both cases, we offered access to subscription video on demand (SVoD) to a random set of households for several weeks and used another random set of households as control group. In both cases, we find that the households induced to binge-watch TV shows are less likely to pay for SVoD after the free trials. Our results suggest that induced binge-watchers deplete the content of interest to them very quickly, which reduces their short-term willingness to pay for SVoD. We also show that carefully crafted recommendation reminders aimed at widening the preferences of consumers for content offset the negative effect of binge-watching and lessen the concerns of binge-watchers with lack of content refresh, and thus these reminders may help content providers manage supply costs, which may otherwise become prohibitive with frequent updates to SVoD catalogs.


  • Effect of Friends’ Churn on Consumer Behavior in Mobile Networks

    by Pedro Ferreira, Rahul Telang, Miguel Godinho De Matos

    We study how consumers decide which tariff plan to choose and whether to churn when their friends churn in the mobile industry. We develop a theoretical model showing conditions under which users remain with their carrier and conditions under which they churn when their friends do. We then use a large and rich anonymized longitudinal panel of call detailed records to characterize the consumers’ path to death with unprecedented level of detail. We explore the structure of the network inferred from these data to derive instruments for friends’ churn, which is typically endogenous in network settings. This allows us to econometrically identify the effect of peer influence in our setting. On average, we find that each additional friend that churns increases the monthly churn rate by 0.06 percent. The observed monthly churn rate across our dataset is 2.15 percent. We also find that firms introducing the pre-paid tariff plans that charge the same price to call users inside and outside the carrier help retain consumers that would otherwise churn. In our setting, without this tariff plan the monthly churn rate could have been as high as 8.09 percent. We perform a number of robustness checks, in particular to how we define friends in the social graph, and show that our results remain unchanged. Our paper shows that the traditional definition of customer lifetime value underestimates the value of consumers and, in particular, that of consumers with more friends due to the effect of contagious churn and, therefore, managers should actively take into account the structure of the social network when prioritizing whom to target during retention campaigns.


  • Target the Ego or Target the Group: Evidence from a Randomized Experiment in Proactive Churn Management

    by Miguel Godinho de Matos, Ferreira Pedro, Belo Rodrigo

    We propose a new strategy for proactive churn management that actively uses social network information to retain consumers.