Likert versus Q-approaches in survey methodologies: discrepancies in results with same respondents

Surveys are constantly used within forest policy processes to obtain information regarding the opinions of stakeholders and the public. Survey outcomes are dependent on what questions are asked, how the answers are structured and analyzed and how interpretations and conclusions are aligned. The most common forms of surveys can be grouped into two categories: the Likert attitude survey, which allows the respondent to freely select an integer value between a minimum and maximum (typically from 1 to 5 or 7), and the Q sort survey, which requires the respondent to sort and classify the importance of questions based on a predefined standard (typically pyramid-shape distribution from disagree to agree). Using similar Likert and Q sort surveys on the same group of respondents in the case of a regional (sub-national) forest policy process, this paper investigates the variability of obtained responses and determines if the choice of survey method would impact the information and potential interpretations provided to the policy makers. The results indicate that survey outcomes may indeed be sensitive to methodology, although respondents would be consistent. When informing policies, it is advisable to carefully consider the purpose and the conduct of the survey, and to make interpretations with prudence.