Gaugris, J.Y.van Rooyen, M.W.2007-07-052007-07-052006Gaugris JY, van Rooyen MW. 2006. Questionnaires do not work! A comparison of methods used to evaluate the structure of buildings and wood used in rural households, South Africa. Ethnobotany Research & Applications 4:119-132.1547-3465http://hdl.handle.net/10125/244The level of hardwood utilisation for house building was evaluated in a rural community of Maputaland, KwaZulu-Natal, South Africa. A full inventory of 42 households in that community was conducted, followed by a questionnaire survey coupled with a partial inventory of the same households. It was expected that the questionnaire design would be greatly improved by the prior survey, and that similar quantitative results could be obtained. The results show that despite a careful design, the questionnaires and coupled partial inventories provided significantly different results, thus placing considerable doubts on any research solely based on questionnaire results without proper ground proofing. The reasons for such differences are unclear. The main advantage of the questionnaire survey resided in the qualitative insight it offered for the analysis of the data.enethnobotanyquestionnairesresearch methodswood constructionrural housingSouth Africahardwoodhousehold surveysinterviewssamplingrural communityrenewable resourcesforest resourcesaccuracyArticle