Conservationists measuring noncompliance with rules about the exploitation of natural resources often need to ask sensitive questions. However, respondents can introduce bias through distorting their answers to direct questions, due to social norms and/or the risk of legal sanctions. Specialized Questioning Techniques (SQTs) are often a more suitable approach to counteracting respondent bias, as they encourage honest answering by protecting respondent’s privacy by design. This study aims to provide a complete overview of the main SQTs, as well as about their most recent advances.We performed a scoping review of existing SQTs, starting from those covered in Nuno and St. John (2015). We included techniques that have never been explained to conservationists before, and/or that were invented after 2015, or which had advanced and improved since Nuno and St. John (2015).Our review identified 9 different broad types of SQTs, defined according to their practical implementation. We found 18 new versions of the randomized response technique, the unmatched count technique, the item sum technique, the triangular model and the crosswise model. We also discuss endorsement experiments and the ballot box method, which have not been covered in a review for the benefit of conservationists. Finally, we found four new SQTs: the parallel model, the pair method, the list method and the person count technique. Each technique is explained with real, or hypothetical, examples and discussed in its advantages and limitations.SQTs have undergone an impressive development over the last few years, and many different techniques are available to elicit sensitive behaviors with implications for conservation. This research offers a summary to conservationists and practitioners who want to understand SQTs and integrate them in their work, which should be a priority for those studying sensitive behaviors with implications for conservation.
On the validity of non-randomized response techniques: an experimental comparison of the crosswise model and the triangular model
