338 www.frontiersinecology.org © The Ecological Society of America I n this article we demonstrate the use of a primary research paper as an assessment tool, helping to determine how well students understand a biological concept and their abil- ity to interpret statistics. The article by Willingham (p 309–313) investigates the effects of temperature in con- junction with endocrine disrupting chemicals (EDCs) on sex determination in turtles. We have designed instruction and assessments on the subject of sex determination, a topic that is conceptually uncomplicated for most students, and statistical analyses which present them with some chal- lenges. We have made the assumption that students have learned about mechanisms of sex determination among ani- mal taxa, know how to develop and test hypotheses, and have a basic understanding of natural selection and fitness.  Student goals • Apply understanding of sex determination to the con- sequences of altered sex ratios in animal populations. • Demonstrate understanding of statistical testing and skills in interpreting data used in a research paper.  Instructor goals • Use primary literature as a source of information about biological topics and as an assessment tool. • Implement an active learning strategy to help students understand the concept of statistical testing and signif- icance.  Engage – content Begin the class with a question for students to discuss in their groups: “In some coastal areas, well-meaning indi- viduals dig up eggs laid by sea turtles on beaches and re- bury them further inland where the eggs are better pro- tected. What impact(s) do you think this has on the sexual development of these turtles?” After selected groups report out, the instructor summa- rizes the discussion, adding information about sex deter- mination and EDCs. The topic of EDCs is of interest to students, as the mechanisms and ubiquity of the effects are easily understood and are of personal relevance. Although controversy exists regarding the links between endocrine disruptors and negative impacts on human health, it is evident that these compounds are present on a global scale, with high levels occurring in the blood or body fats of humans and wildlife. This introduction leads to an exploration of the use of statistics in these types of studies.  Explore – statistics and data Two objectives guide students’ exploration of statistics: (1) connect data interpretation and confidence level with statistical testing, and (2) connect statistical analy- ses of data with support or rejection of hypotheses. The instructor polls the class, using computer-based personal response systems, “clickers” (Brewer 2004), or hand-held cards to provide real-time displays of responses to questions like the following, using five choices (eg 100%, 95–85%, 75–65%, 55–45%, < 40%): “What level of confidence does an engineer need to have in a new material for building bridges for public use before the bridges are actually built with that material?” Students usually say 100%. After a discussion that perfection is never possible, groups of students decide what level of confidence they think is acceptable for these and similar items. These types of questions lead naturally to a discussion of the effects of sample size and data variability on confidence in results. Sample size and confidence levels Students then discuss the following scenario to explore determination of sample size: “Suppose you are the head of a drug-testing team. You have a pool of 10 000 people on whom to test the drug. What sample size of people will you use – 10, 100, or 1000? You may assume that one individual by chance alone responds unpredictably to the drug. Explain your answer in terms of your confidence that the drug effects on all individuals measured are truly representative of the drug.” In groups, students calculate the impact of one anom- alous outcome within a sample of 10, 100, or 1000 and relate their solutions to thinking about sample size and confidence levels. Why would they/would they not use a smaller or larger sample size? Students report out answers. P-values The next hurdle for students is in understanding what P- values mean in relation to hypothesis testing. Most stu- dents view hypotheses as absolutes – ie right or wrong – and have difficulty understanding statistical significance. The instructor builds on the previous activity by explain- ing the meaning of a P-value and that 0.05, which is equivalent to a 95% confidence level, is a value tradition- ally used to indicate actual rather than chance effects of treatments if the null hypothesis is true.  Explain – analysis Students work in groups to explore the concept of statistical significance using P-values. Groups work with one of two datasets, representing the occurrence of cricket frogs by gonadal sex in relation to time period and/or geographic Determining confidence: sex and statistics Terry L Derting 1 , Diane Ebert-May 2 , Janet Hodder 3 , and Everett P Weber 2 PATHWAYS TO SCIENTIFIC TEACHING 1 Murray State University, 2 Michigan State University, 3 University of Oregon