Seeing Green: Do Visual Representations Matter in Economic Games

We use a within-subjects experimental design. Participants receive the control condition in stages 1 and 3 and the treatment in stage 2.
Control condition: Participants are told how many acres of forested land they have remaining and are then asked to make a harvesting decision.
Treatment condition: Participants are told how many acres of forested land they have remaining and are provided with a visual representation of their remaining forested land and are then asked to make a harvesting decision

2026-07-21

2026-07-27

1) total number of acres harvested by each individual in each of the three stages
2) total number of forested acres remaining for each individual at the end of each stage
3) the number of acres harvested by each individual during each round of each stage

The study proceeds as follows:
1) Participants choose this study from the list of available studies on CloudResearch Connect and click a link that directs them to the online survey experiment hosted on Qualtrics.
2) After providing consent, participants encounter an attention-check question.
3) Participants who pass the attention check, receive information about how the forest management game is played.
4) Participants are presented with two comprehension check multiple choice questions to assess whether they understand how the game works.
5) Participants who answer each of the comprehension-check questions correctly within the three attempts allowed, go on to participate in a practice stage.
6) Participants then encounter Stages 1 - 3 of the game.
7) Finally, participants complete a brief survey about their attitudes, experiences, and sociodemographics.

Game Play:
The forest management game is played in 3 stages of 5 rounds each. Participants begin each stage with 40 acres of forested land. In each round, participants decide how many acres of their forested land to harvest. They receive 9 points per acre harvested. At the end of rounds 1 - 4, forested land regrows at a rate of 1 acre per every 4 acres of forested land remaining. A full 4 acres must be present in order to regrow an acre of forest (e.g., if 11 acres of land remain as forests, only 2 acres of forests will regrow). Forests do not regrow after participants make their Round 5 decision. At the start of each round, participants are told how many acres of forested land they have remaining and are asked to report how many acres they would like to harvest.
In Stage 2, participants receive a visual representation of their forested land in addition to being told the number of forested acres remaining.
One of the three stages played is chosen at random to determine the participant’s bonus payment. The stage is referred to as the payment stage.
At the end of each stage, participants are provided with information about the total acres harvested and the total number of forested acres remaining. They are also told the total number of points they would earn and total number of tree-plantings that would occur should the current stage be chosen as the payment stage.

Survey:
After completing the land management game, participants are asked to complete a survey about their environmental attitudes, exposure to nature, and sociodemographic information. They are also asked to report whether the visual representation in stage 2 affected their understanding of the game play or their harvesting behavior, whether they believe games such as this can help us understand real world conservation behavior, and whether their own behavior in the game is a realistic representation of how they might behave if they were a forest landowner.
At the end of the survey, participants are informed of the stage that was randomly selected for payment, the bonus payment they have earned, and the number of tree plantings they have contributed. Payment is made through CloudResearch Connect.

See pre-analysis plan for full detail

There is no random assignment to control or treatment. This is a within-subjects design. All participants receive both the control condition (stage 1 and stage 3) and the treatment condition (stage 2).

NA

No

310 individuals are recruited

Each participant plays 15 rounds of the game 310 * 15 = 4,650 observations in longest format (round-level analysis) 310 * 3 = 930 observations in shorter format (stage-level analysis)

NA

Using data from a pilot study performed in late fall 2024, estimated average conserved forested land is 13 acres with a standard deviation of 13 acres. Assuming differences in means with paired samples, if we want to be able to identify an average treatment effect of approximately 2 acres (effect size of approximately 0.16), we will need approximately 310 participants in our study (α=0.05,power=0.80) (gpower 3.1.9.7).

Visualization Study Pre-Analysis Plan