Correcting Misperceptions about Nuclear Energy: A Survey Information Experiment in the Netherlands

2026-06-03

2026-06-16

Factual belief accuracy regarding nuclear energy, measured as the pre-to-post change in the number of correct answers to factual questions on CO₂ emissions, reactor explosions, cooling towers, and radioactive waste.

1) Factual belief accuracy regarding nuclear energy

A factual belief accuracy index measuring the number of correct answers to four factual questions about nuclear energy. The index ranges from 0 to 4 and includes:

Whether nuclear energy produces CO₂ emissions comparable to fossil fuels during electricity generation.
Whether a nuclear reactor can explode in the same way as an atomic bomb.
Whether cooling towers emit radioactive smoke.
The amount of high-level radioactive waste produced annually by the Borssele nuclear power plant.

The primary outcome is the change in this index between the pre-treatment and post-treatment measurements.

2) Risk perceptions regarding nuclear energy

Agreement with the statement that radioactive waste can be stored and managed in ways designed to prevent harm to people and the environment.
Perceived safety of nuclear energy relative to other energy sources.

The secondary outcomes are measured as changes between pre-treatment and post-treatment responses.

3) Support for nuclear energy expansion

Support for expanding nuclear energy in the Netherlands, measured on a 0–100 scale.

Local acceptance of nuclear facilities

Willingness to accept a nuclear power plant near one's residence, measured through support for a nearby nuclear facility and preferred minimum distance from home.

These outcomes are measured before and after treatment and analysed as exploratory attitudinal outcomes.

This study is an online survey experiment conducted among Dutch residents. Respondents are randomly assigned at the individual level to either a control condition or a treatment condition.

All respondents first complete pre-treatment questions measuring factual beliefs about nuclear energy, broader risk perceptions, and policy preferences. Respondents in the control condition then receive a neutral informational text about electricity generation and nuclear energy. Respondents in the treatment condition receive a text containing corrective factual information addressing common misperceptions about nuclear energy.

After reading the assigned text, respondents complete post-treatment questions measuring the same main outcomes. The study combines individual-level random assignment with a within-person pre-post measurement design.

Not available

Simple individual-level random assignment implemented through the Qualtrics randomizer with equal probability of assignment to treatment and control conditions (50/50 allocation).

Individual respondent.

No

0

100 - 150 individuals

Treatment = 75
Control = 75

Power calculations were conducted for the primary outcome: the change in the factual belief accuracy index from pre- to post-treatment. The index ranges from 0 to 4 and is based on four factual belief items. The unit of analysis is the individual respondent. The primary estimand is the difference in average pre-post change between treatment and control respondents. The study uses individual-level random assignment, with no planned clustering. The expected final sample size is between 100 and 150 respondents, with approximately equal assignment to treatment and control. Using a two-sided significance level of α = .05 and target power of 80%, the minimum detectable effect ranges from approximately 0.57 points with 100 respondents to 0.46 points with 150 respondents, assuming a standard deviation of 1.00 for the change score. Sensitivity analyses using alternative standard deviations suggest that the minimum detectable effect ranges from 0.42 to 0.71 points with 100 respondents, and from 0.35 to 0.58 points with 150 respondents. The expected treatment effect is approximately 0.60 points on the factual belief accuracy index. This is based on anticipated improvements of 5 percentage points for the CO₂ item, 5 percentage points for the reactor-explosion item, 25 percentage points for the cooling-tower item, and 25 percentage points for the radioactive-waste item. Under these assumptions, a sample size between 100 and 150 respondents may be sufficient to detect the expected primary effect, although the study may be underpowered to detect smaller effects. Therefore, statistically insignificant findings will be interpreted cautiously and will not necessarily be taken as evidence of no treatment effect. Secondary outcomes and subgroup analyses will be interpreted as exploratory.