Large-scale Online Experiment Research Articles

Social Information Disclosure in C2C E-commerce: A Natural Experiment

This paper aims to explore the influecne of social information disclosure on the sales rate in the online C2C trading market, and a mediator effect of privacy and a moderator effect of trading experience are considered. We conducted a large-scale online experiment in 80 days on a C2C commodity trading platform in China. Based on the experiment data, the empirical results show that partial social information disclosure would decrease the sales rate while full social information disclosure would increase the sales rate. The mediator effect of privacy would increase the sales rate, but the impact is relatively small compared to the trust. The moderator effect of trading experience would only enhance the negative influence of partial social information disclosure but have no effect on the influence of full social information disclosure. This paper encourages the cooperation between social platforms and trading platforms by introducing the real social relationship into the online trading market.

Computational modelling of attentional selectivity in depression reveals perceptual deficits.

Depression is associated with broad deficits in cognitive control, including in visual selective attention tasks such as the flanker task. Previous computational modelling of depression and flanker task performance showed reduced pre-potent response bias and reduced executive control efficiency in depression. In the current study, we applied two computational models that account for the full dynamics of attentional selectivity. Across three large-scale online experiments (one exploratory experiment followed by two confirmatory - and pre-registered - experiments; total N = 923), we measured attentional selectivity via the flanker task and obtained measures of depression symptomology as well as anhedonia. We then fit two computational models that account for the dynamics of attentional selectivity: The dual-stage two-phase model, and the shrinking spotlight (SSP) model. No behavioural measures were related to depression symptomology or anhedonia. However, a parameter of the SSP model that indexes the strength of perceptual input was consistently negatively associated with the magnitude of depression symptomatology. The findings provide evidence for deficits in perceptual representations in depression. We discuss the implications of this in relation to the hypothesis that perceptual deficits potentially exacerbate control deficits in depression.

The impact of markets on moral reasoning: Evidence from an online experiment

This paper investigates the impact of markets on moral reasoning. Whereas the current literature focuses on morally relevant decisions that arise in markets, little is known about whether the exposition to markets shapes subsequent moral reasoning. To close this gap, we run a large-scale online experiment with 3 conditions: In Baseline, participants make a choice in a moral dilemma. In the other two conditions, participants are exposed to either a Non-market or Market environment, before facing the identical choice in the moral dilemma. We hypothesize that being exposed to Market induces cost-benefit considerations, which translate into modified reasoning in the subsequent moral dilemma. Compared to the baseline distribution, we indeed find a substantial effect in Market. However, similar choices can be observed in Non-market. We discuss potential explanations for these results, and suggest avenues for future research.

Perceived-Value-driven Optimization of Energy Consumption in Smart Homes

Residential energy consumption has been rising rapidly during the last few decades. Several research efforts have been made to reduce residential energy consumption, including demand response and smart residential environments. However, recent research has shown that these approaches may actually cause an increase in the overall consumption, due to the complex psychological processes that occur when human users interact with these energy management systems. In this article, using an interdisciplinary approach, we introduce a perceived-value driven framework for energy management in smart residential environments that considers how users perceive values of different appliances and how the use of some appliances are contingent on the use of others. We define a perceived-value user utility used as an Integer Linear Programming (ILP) problem. We show that the problem is NP-Hard and provide a heuristic method called COndensed DependencY (CODY). We validate our results using synthetic and real datasets, large-scale online experiments, and a real-field experiment at the Missouri University of Science and Technology Solar Village. Simulation results show that our approach achieves near optimal performance and significantly outperforms previously proposed solutions. Results from our online and real-field experiments also show that users largely prefer our solution compared to a previous approach.

Understanding Tax Policy: How Do People Reason?

I study how people understand, reason, and learn about tax policy. The goal is to uncover the mental models that people use to think about income and estate taxes. To that end, I run large-scale online surveys and experiments on representative U.S. samples to elicit not only respondents' factual knowledge about tax policy and the income or wealth distributions, but also their understanding of the mechanisms of tax policy and their reasoning about it. The detailed survey questions are designed to address the three main factors emphasized in our core tax model that can shape support for or opposition to taxes: efficiency effects, distributional implications, and fairness considerations. But they also elicit broader concerns that could influence policy views, such as misperceptions, views of government, perceived spillovers from taxes, and views on how tax revenues are or should be spent. I decompose policy views into the various underlying factors and find that support for tax policy is most strongly correlated with views on the benefits of redistribution and fairness, as well as with views of the government. concerns play a more minor role. These correlational patterns are confirmed by the experimental approach, which shows people instructional videos that explain the workings and consequences of one of the aspects of tax policy (the and the Efficiency treatments) or that bring the two together and focus on the trade-off (the treatment). The Redistribution treatment and Economist treatments significantly increase support for more progressive taxes. I also find that there are partisan divergences not just in the final policy views, but also at every step of the reasoning about the underlying mechanisms of taxes, and most starkly on the fairness considerations. Institutional subscribers to the NBER working paper series, and residents of developing countries may download this paper without additional charge at www.nber.org.

Institutions, Opportunism and Prosocial Behavior: Some Experimental Evidence

Formal or informal institutions have long been adopted by societies to protect against opportunistic behavior. However, we know very little about how these institutions are chosen and their impact on behavior. We experimentally investigate the demand for different levels of institutions that provide low to high levels of insurance and its subsequent impact on prosocial behavior. We conduct a large-scale online experiment where we add the possibility of purchasing insurance to safeguard against low reciprocity to the standard trust game. We compare two different mechanisms, the private (purchase) and the social (voting) choice of institutions. Whether voted or purchased, we find that there is demand for institutions in low trustworthiness groups, while high trustworthiness groups always demand lower levels of institutions. Lower levels of institutions are demanded when those who can benefit from opportunistic behavior, i.e. low trustworthiness individuals, can also vote for them. Importantly, the presence of insurance crowds out civic spirit even when subjects can choose the no insurance option: trustworthiness when formal institutions are available is lower than in their absence.

Institutions, Opportunism and Prosocial Behavior: Some Experimental Evidence

Formal or informal institutions have long been adopted by societies to protect against opportunistic behavior. However, we know very little about how these institutions are chosen and their impact on behavior. We experimentally investigate the demand for different levels of institutions that provide low to high levels of insurance and its subsequent impact on prosocial behavior. We conduct a large-scale online experiment where we add the possibility of purchasing insurance to safeguard against low reciprocity to the standard trust game. We compare two different mechanisms, the private (purchase) and the social (voting) choice of institutions. Whether voted or purchased, we find that there is demand for institutions in low trustworthiness groups, while high trustworthiness groups always demand lower levels of institutions. Lower levels of institutions are demanded when those who can benefit from opportunistic behavior, i.e. low trustworthiness individuals, can also vote for them. Importantly, the presence of insurance crowds out civic spirit even when subjects can choose the no insurance option: trustworthiness when formal institutions are available is lower than in their absence.

The effect of experts’ and laypeople’s forecasts on others’ stock market forecasts

Abstract With a large-scale online experiment with 1593 participants from the U.S. and the U.K. we explore whether and how people working in the finance industry and laypeople from the general population are influenced by information on other people’s forecasts when making forecasts on the future development of two indices and two stocks. We find that (i) laypeople’s forecasts are strongly influenced by information they get on other subjects’ forecasts, while financial professionals are much less influenced by information signals; (ii) signals by financial professionals influence all subject groups more than forecasts by laypeople; (iii) we observe a home bias in all subject groups, which can be mitigated by information signals; (iv) all subject groups expect lower forecast errors for financial professionals than for laypeople, hence we find evidence for trust in experts.

Lying and reciprocity

We extend the literature on the determinants of lying to the impact of reciprocity in strategic interactions. We study a theoretical model with reciprocity by assuming that a player's degree of altruism depends on her perception of the other player's altruism towards herself. Specifically, we consider a sequential two-player-contest and vary the second mover's degree of information on the first movers lying behavior. This yields predictions on the second mover's behavior which we study in a largescale online-experiment and the laboratory. Surprisingly, the second mover's lying propensity does not depend on her information on the first mover's lying behavior. To investigate whether this is specific to lying, we consider an additional treatment with the same payoff structure where subjects can increase their chances of winning by simple statements without lying. As we do find evidence for reciprocity there, the lack of reciprocity cannot solely be attributed to the contest structure.

Enhancing central bank communications using simple and relatable information

Central bankers and the central banking literature are increasingly attuned to the importance of communications as a policy tool. However, less is known about how central bank communications should be drafted for maximal impact. Our paper contributes new insights in this regard. Using a large-scale online experiment with a sample representative of the UK population, the paper documents the communicative techniques that increase public comprehension and trust in monetary and macroeconomic policy messages. Key findings include that the simplification of language increases public comprehension more than the inclusion of visuals, and that public comprehension can be improved by making monetary policy messages relatable to people's lives. Relatable content also increases the public's trust in central bank communications, and improves people's perceptions of the central bank. Our findings shed light on how central banks can improve communication with the public at a time when trust in public institutions has fallen, while the responsibilities delegated to central banks have increased.

Some remarks on the comparative experiment as a method in assessing populist political communication in Europe

In this paper fundamental information concerning the designing and conducting of comparative experiments, as methods which are able to be employed in the field of media studies, will be successively delineated. On the basis of the large-scale online experiment conducted in 15 countries in 2017, the assessment of the populist message impact as well as methodological challenges to such a project will be presented. We discuss challenges and lessons learnt from this type of research design.

Personalized Reason Generation for Explainable Song Recommendation

Personalized recommendation has received a lot of attention as a highly practical research topic. However, existing recommender systems provide the recommendations with a generic statement such as “Customers who bought this item also bought…”. Explainable recommendation, which makes a user aware of why such items are recommended, is in demand. The goal of our research is to make the users feel as if they are receiving recommendations from their friends. To this end, we formulate a new challenging problem called personalized reason generation for explainable recommendation for songs in conversation applications and propose a solution that generates a natural language explanation of the reason for recommending a song to that particular user. For example, if the user is a student, our method can generate an output such as “Campus radio plays this song at noon every day, and I think it sounds wonderful,” which the student may find easy to relate to. In the offline experiments, through manual assessments, the gain of our method is statistically significant on the relevance to songs and personalization to users comparing with baselines. Large-scale online experiments show that our method outperforms manually selected reasons by 8.2% in terms of click-through rate. Evaluation results indicate that our generated reasons are relevant to songs and personalized to users, and they attract users to click the recommendations.

Understanding node-link and matrix visualizations of networks: A large-scale online experiment

AbstractWe investigated human understanding of different network visualizations in a large-scale online experiment. Three types of network visualizations were examined: node-link and two different sorting variants of matrix representations on a representative social network of either 20 or 50 nodes. Understanding of the network was quantified using task time and accuracy metrics on questions that were derived from an established task taxonomy. The sample size in our experiment was more than an order of magnitude larger (N = 600) than in previous research, leading to high statistical power and thus more precise estimation of detailed effects. Specifically, high statistical power allowed us to consider modern interaction capabilities as part of the evaluated visualizations, and to evaluate overall learning rates as well as ambient (implicit) learning. Findings indicate that participant understanding was best for the node-link visualization, with higher accuracy and faster task times than the two matrix visualizations. Analysis of participant learning indicated a large initial difference in task time between the node-link and matrix visualizations, with matrix performance steadily approaching that of the node-link visualization over the course of the experiment. This research is reproducible as the web-based module and results have been made available at: https://osf.io/qct84/.

Effective strategies for responding to rumors about risks: The case of radiation-contaminated food in South Korea

This experimental study explores how governments should respond to rumors about national-level risk issues. Informed by research in rumor psychology and risk/crisis communication, it investigates whether type of rumor and rumor response strategy have main and interaction effects on reducing rumor beliefs and intention to disseminate rumor. The two featured rumor types are the bogie rumor, which highlights feared outcomes, and the wedge rumor, which aims to reinforce differences between rival groups. Derived from Situational Crisis Communication Theory, the three response strategies examined are refuting the rumor, denying it, and attacking its source. Data were drawn from part of a large-scale online experiment, and the sample of the analysis was 942 South Korean adults. The experiment had a between-subjects design of 2 rumor type (wedge vs. bogie) x 3 government response strategies (refutation, denial, attack the attacker). Results show that all three rumor response strategies significantly reduced rumor beliefs, but only the refutation strategy significantly reduced intention to disseminate the rumor. Rumor type (bogie) and response strategies (refutation) had main, but not interaction, effects on reduction of intention to disseminate the rumor.

Lower socioeconomic status is associated with higher intended consumption from oversized portions of unhealthy food

Socioeconomic status is one of the strongest predictors of obesity, and of living in deprived neighbourhoods with unhealthy food environments. Little is known, however, about the psychological processes that translate features of such environments into socioeconomic differences in eating behaviour. One important feature of unhealthy food environments is the prevalence of oversized portions of unhealthy food. The present study tested whether individuals with lower socioeconomic status intend to consume more from large portions than those with higher socioeconomic status, and examined the psychological processes underlying this effect. A large-scale online experiment was conducted in which participants (N = 511) indicated how much they would eat from small and large portions of healthy and unhealthy snacks. The mediating effects of trait impulsivity and perceptions of how much was considered appropriate to eat were also assessed. Participants with lower socioeconomic status intended to eat more from the large portions than from the small portions of the unhealthy snacks, which would equate to a potential 15–22% increase in energy intake. These effects were partially mediated by trait impulsivity and perceptions of how much is appropriate to eat. These findings point to a significant health burden of low socioeconomic status: when exposed to unhealthy food environments, specific psychological processes might increase the amount of unhealthy food those with lower socioeconomic intend to consume. This study critically informs the emerging understanding of the psychology of socioeconomic status and eating behaviour.

Unfounded beliefs among teachers: The interactive role of rationality priming and cognitive ability

SUMMARYPrevious research suggests that unfounded beliefs (UB)—such as conspiracist beliefs and beliefs in the supernatural—stem from similar cognitive and motivational mechanisms. More specifically, it has been demonstrated that cognitive ability is negatively associated with UB but only among individuals who value epistemic rationality. The present study goes beyond previous correlational studies by examining whether the negative association between cognitive ability and UB can be strengthened through a subtle rationality prime. In a large scale online experiment (N = 762 French teachers), we demonstrate that priming rationality (vs. control) does enhance the negative relationship between cognitive ability and adherence to supernatural beliefs, as well as conspiracy mentality (d = 0.2). This effect was not obtained for illusory pattern perception. This study's usefulness as a “proof of concept” for future interventions aimed at reducing UB prevalence among the general public is discussed.

Fake images: The effects of source, intermediary, and digital media literacy on contextual assessment of image credibility online

Fake or manipulated images propagated through the Web and social media have the capacity to deceive, emotionally distress, and influence public opinions and actions. Yet few studies have examined how individuals evaluate the authenticity of images that accompany online stories. This article details a 6-batch large-scale online experiment using Amazon Mechanical Turk that probes how people evaluate image credibility across online platforms. In each batch, participants were randomly assigned to 1 of 28 news-source mockups featuring a forged image, and they evaluated the credibility of the images based on several features. We found that participants’ Internet skills, photo-editing experience, and social media use were significant predictors of image credibility evaluation, while most social and heuristic cues of online credibility (e.g. source trustworthiness, bandwagon, intermediary trustworthiness) had no significant impact. Viewers’ attitude toward a depicted issue also positively influenced their credibility evaluation.

Fake Images: The Effects of Source, Intermediary, and Digital Media Literacy on Contextual Assessment of Image Credibility Online

Fake or manipulated images propagated through the Web and social media have the capacity to deceive, emotionally distress, and influence public opinions and actions. Yet few studies have examined how individuals evaluate the authenticity of images that accompany online stories. This paper details a 6-batch large-scale online experiment using Amazon Mechanical Turk that probes how people evaluate image credibility across online platforms. In each batch, participants were randomly assigned to one of 28 news-source mockups featuring a forged image, and they evaluated the credibility of the images based on several features. We found that participants’ internet skills, photo-editing experience, and social media use were significant predictors of image credibility evaluation, while most social and heuristic cues of online credibility (e.g., source trustworthiness, bandwagon, intermediary trustworthiness) had no significant impact. Viewers’ attitude towards a depicted issue also positively influenced their credibility evaluation.

Behavioural realism and the activation of aggressive concepts in violent video games

A common argument in the violent video game (VVG) literature is that the greater the realism of a game, the more it activates aggressive concepts, and the greater antisocial effects it will have on its players.Several experiments have therefore looked into whether the graphical realism of VVGs might influence their effects. These experiments have returned mixed results. However, there are other ways that a VVG can be realistic besides looking like the real world. More specifically, things in VVGs can not only look realistic, they can also behave realistically. It may be the case that this kind of realism leads to increases in the activation of aggressive concepts, rather than increases in graphical realism.In this paper, we therefore present two large-scale online experiments (n = 898 and n = 1880) which investigate the effects of two different manipulations of behavioural realism on the activation of aggressive concepts in VVGs. In neither experiment did increasing realism increase the activation of aggressive concepts.Realism is often described as increasing the effects of VVGs. These results contradict this perspective, and instead suggest that realism may not lead to increases in aggression-related variables.

The Usefulness of Financial Accounting Information: Evidence from the Field

We examine how investment professionals assess the usefulness of financial accounting information depending on their information acquisition objectives and preparers’ earnings management incentives. We conduct a survey experiment based on face-to-face interviews with investment professionals and document two main results. First, we find that, compared with investment professionals assigned a firm valuation objective, those assigned a managerial performance evaluation objective assess accounting information as significantly less useful. Second, we find no systematic evidence that preparers’ earnings management incentives negatively affect investment professionals’ assessments of accounting information usefulness. To elucidate this second finding, we conduct a large-scale follow-up online experiment. Our results continue to offer no support for the effect of earnings management incentives on investment professionals’ assessments of accounting information usefulness, irrespective of preparers’ corporate governance quality. Instead, we find that poor corporate governance, by itself, reduces the usefulness of accounting information to investment professionals.