Discrete choice experiments (DCE) are one of the main methods for the valuation of non-market environmental goods. However, concerns regarding the validity of choice responses obtained in such surveys remain, particularly in surveys dealing with environmental goods remote from and unfamiliar to respondents. This study assesses behavioural determinants of preferences for conservation benefits of a marine protected area on the Dogger Bank, a shallow sandbank in the southern North Sea in an attempt to assess construct validity of survey responses. The Theory of Planned Behavior (TPB) and the Norm Activation Model (NAM) are employed to empirically measure constructs that predict stated choices. The study finds that identified protest respondents score significantly lower on most TPB and NAM components than non-protesters. Results further show that components of both the TPB and the NAM robustly predict choice behaviour. The inclusion of the TPB components improves the predictive power of the estimation model more than the NAM components. In an additional latent class logit model, TPB and NAM components plausibly explain different patterns of WTP for conservation benefits of an offshore marine protected area. These findings support construct validity of stated choice data regarding the valuation of remote and unfamiliar environmental goods.