Existing Positions, Information Acquisition, and Investment Decisions

Participants make an incentivised choice between two investments that differ in their private payoff and environmental consequences. Three between-subjects conditions, labelled Treatments A, B and C, vary whether information about the environmental consequences is initially available and whether one investment is established as the participant's existing position. The design therefore allows us to examine whether an existing position affects optional information acquisition and subsequent choice. Eye-tracking data are collected as an exploratory process measure.

2026-09-28

2026-10-27

Information acquisition, measured as whether the participant accesses the optional information about the environmental consequences before submitting the decision. Investment choice, measured as which of the two investments the participant holds at the end of the task.

The focal contrast is Treatment C against Treatment B. Two focal hypotheses carry the confirmatory claim. H2 concerns information acquisition in Treatment C against Treatment B. H3 concerns the final investment choice in Treatment C against Treatment B, restricted to a pre-specified subsample. Each carries one pre-specified one-sided two-sample proportion test at alpha = .05, and both are reported whatever the outcome. The Holm-Bonferroni result over H2 and H3 at a family-wise alpha of .05 is reported alongside as a pre-specified supplementary analysis. H1, a directional benchmark on the investment choice between the treatment in which the information is initially available and the treatment in which it is not, is pre-registered and tested one-sided at alpha = .05. It is reported separately from the two focal hypotheses and is not part of the supplementary Holm analysis. Both outcomes are recorded automatically by the experimental software. Robustness analyses are two-sided tests and re-estimation excluding flagged participants; any further analyses are exploratory.

Visual attention to the two payoff dimensions on the decision screen, measured by eye tracking and analysed exploratorily; the position manipulation check; post-decision mechanism items; and a control item on awareness of the gaze recording.

The gaze layer is specified in advance and is exploratory throughout. It carries no confirmatory hypothesis and does not affect the behavioural hypotheses. Its focus is whether an existing position changes how participants allocate attention on the decision screen before the decision is submitted, in particular between the two payoff dimensions. Gaze later in the decision screen is examined as well. Gaze is recorded over the full screen, so further process measures may be added, and all gaze results are labelled exploratory. The position manipulation check comprises six seven-point items on how participants experienced the position and is reported descriptively. The mechanism items and all further gaze analyses are exploratory.

A three-condition between-subjects laboratory experiment conducted at the Mainz Behavioral and Experimental Laboratory (MABELLA). Participants make one incentivised choice between two investments that differ in private payoff and an associated carbon-offset contribution. Across conditions, we manipulate the availability of information about the environmental consequences and whether one option constitutes an existing position. Random assignment occurs at the individual level within experimental sessions.

Not available

Random assignment by the experimental software at session entry, with equal probabilities across the three treatments. Assignment is at the individual level within sessions, and treatments are mixed within the same session. Demographic characteristics are summarised by treatment and reported descriptively.

Individual participant

No

Not applicable, randomisation is at the individual level.

300 participants in the main study, one incentivised decision each. A completed pretest of 54 recordings across three sessions is reported separately and is not pooled with main-study data.

Approximately 100 analysable participants in each of the three treatments. Recruitment carries a margin for no-shows. The gaze subsample is not recruited separately.

With approximately 100 analysable participants per arm, a one-sided two-sample test of proportions at an alpha of 5% and a power of 80% detects a difference of 17.5 percentage points or more in the information-acquisition outcome, for any rate in the comparison treatment. The choice outcome is analysed on a restricted sample of approximately 90 per arm, where the corresponding value is 18.4 points. For the supplementary Holm-Bonferroni analysis over H2 and H3 the values are 19.7 and 20.8 points.