Recent Survey Experiments in the Science and Engineering Indicators 2016

Sunday, February 14, 2016: 1:00 PM-2:30 PM
Marriott Balcony B (Marriott Wardman Park)
Peter Muhlberger, National Center for Science and Engineering Statistics, Arlington, VA
Science and Engineering Indicators, a publication of the National Science Foundation’s (NSF’s) National Science Board, funds and reports survey research regarding public knowledge of science.  Questions are asked on the General Social Survey (GSS).  Past surveys have found that Americans do poorly in correctly answering a question about whether humans developed from earlier species of animals.  This raises the issue of whether a differently worded question could better capture knowledge of evolution by not running counter to personal beliefs.  In addition, scholars have questioned whether the battery of GSS questions are asked in ways that best capture public knowledge—particularly, whether true / false questions might better be offered as forced-choice questions and whether respondents give more reliable answers if not allowed a ‘do not know’ option.  Researchers provide conflicting theories and findings about best practices to achieve reliable responses.  This presentation presents findings from survey experiments conducted to answer these questions.  Findings include that there is a better way to ask Americans about evolution and that current question practices do about as well as possible in capturing factual knowledge of science.