Submitting results...

Success! Your study has concluded.

Error. Your results have somehow failed to submit. Please email decisionresearchlab@outlook.com referencing the following information and we will look into the issue:

Participant Id:

Survey id:

Error:

What's the effect size?

On the next page we're going to show you an extended abstract for a paper. Please read it as if you were an editor for a scientific journal, deciding on whether to accept the paper. After you read the abstract we'll ask you a few questions.

Note: If you complete this study, we will donate a set of personal protective equipment (PPE) for a UN refugee health worker.

Click here to confirm you have not completed this study before.

Please confirm you understand the scenario to proceed.

Click here to confirm you understand the scenario

Extended abstract:
Video games and aggressive behavior in the laboratory

Note: We know this looks long, but this is the only reading in the study. The rest of the pages are very short.

Entertainment media affects our lives. What behaviors children and adults consider appropriate comes, in part, from the lessons learned from television and the movies. Here we use a randomized control trial in a laboratory setting to test the impact of playing violent video games on aggressive behavior. We recruited a sample of 800 undergraduates from a large Midwestern introductory psychology course to participate in a lab experiment where their aggression was assessed after playing a violent or non-violent single player video game.

Participants were randomly assigned to treatment (violent game) or control (non-violent game) conditions, with 400 participants in each condition. After playing the game for a total of 90 minutes they were placed in unrelated situations where they could anonymously be more or less aggressive towards another person. Our main finding is that participants who played violent video games behaved more aggressively (M = 6.81, SD = 0.51) compared to those who played non-violent video games (M = 6.65, SD = 0.53), t(798) = 4.35, p < .001 (Fig. 1).

Specifically, after playing the game participants were told that someone occupying a neighboring cubicle would compete with them to press a button first in response to a stimulus, and the loser would be subjected to a noise blast, though the neighboring cubicle was in fact empty, and the participant won or lost rounds at random. The intensity and duration of the noise blast was set by the participant ahead of time. When they won, they heard the noise blast in the neighboring cubicle. When they lost, they were subjected to a noise blast of a random intensity and length. We measured their aggressiveness by the length of the noise blast that participants chose to deliver to their opponent when they won the reaction time task.

In the short term, playing a violent video game appears to affect aggression by priming aggressive thoughts. If repeated exposure to violent video games does indeed lead to the creation and heightened accessibility of a variety of aggressive knowledge structures, the consequent changes in everyday social interactions may also lead to consistent increases in aggressive affect.

Figure 1. Participants who played violent video games had a higher aggressiveness score (as measured by the length of noise blasts they delivered to their opponents, in log milliseconds) compared to those who played non-violent video games. Error bars show 1 standard error above and below the mean for each condition. Colored points show all of the individual responses in each condition.

Please evaluate this extended abstract to the best of your ability on the following dimensions.

This question is required.

Appeal of this work to a broad interdisciplinary audience:

1 (Not at all appealing)

5 (Very appealing)

This question is required.

How sufficient is the experiment's sample size?

1 (Completely insufficient)

5 (Completely sufficient)

This question is required.

Overall recommendation of abstract for publication:

1 (Strong accept)

2 (Accept)

3 (Possible reject)

4 (Reject)

5 (Strong reject)

Please answer the following question to the best of your ability

What is the probability that a randomly selected member of the treatment group (someone who played a violent video game) had a higher aggressiveness score than a randomly selected member of a control group (someone who played a non-violent video game)?

(A 50% probability indicates no difference in outcomes between the treatment and control groups, on average.)

A question about data from the previous page

Warning: do not press the back button, that will exit the experiment

What kind of error bars were shown in the plot on the previous page?

This question is required.

1 standard error above the mean, and 1 standard error below the mean

1 standard deviation above the mean, and 1 standard deviation below the mean

2 standard errors above the mean, and 2 standard errors below the mean

2 standard deviations above the mean, and 2 standard deviations below the mean

What did the error bars on the previous page capture conceptually?

This question is required.

Uncertainty in estimating the average in each condition

Variation in individual outcomes in each condition

What is the effect size?

Trial 1 of 1000

Suppose you're looking at the results of an experiment involving two groups of people: 74 people who received a treatment and 75 people who were in a control condition. The outcomes in each condition follow a bell curve (or normal distribution), as you might see for people’s heights or weights. The accompanying plot shows the estimated average outcome for each group with error bars showing one standard error above and below the average in each condition.

What is your best estimate of the probability that a randomly selected person in the treatment group has a higher outcome than a randomly selected person in the control group? (A 50% probability indicates no difference in outcomes between the treatment and control groups, on average.)

Estimate:

Answer:

Here’s a different view of the same exact data and estimates. The error bars still show the same standard errors around the estimated average in each group, but appear smaller because we have increased the range on the y axis. We’ve done so in order to show all of the individual outcomes.

Link to paper.

Return to homepage.

What you are seeing is about to change

The error bars will still show one standard error around the estimated average in each group. In addition, we will now show all of the individual outcomes.

I understand

What you are seeing is about to change

The error bars will still show one standard error around the estimated average in each group, but we will no longer show individual outcomes.

I understand

A few final questions about you and your experience

1. What sources of formal or informal training or education have you had in research, research methods, or statistics? Select all that apply.

Undergraduate coursework

Master's program in a data science related field

PhD program in a data science related field

Postgraduate coursework (e.g., during internship, fellowship, etc...)

Continuing education courses

Self-instruction via textbooks or peer-reviewed literature

Other

2. How comfortable are you with understanding the results of randomized experiments, including reading graphs and plots?

Not at all

Somewhat

Moderately

Very

Extremely

3. Have you ever done any of the following? Select all that apply.

Read the results of a randomized experiment in a peer-reviewed journal article