One-Sample Hypothesis Testing Let's assume that the United States population is evenly split; so exactly half of the U.S. are men and the other half of the U.S. are women. In SPSS, the variable sex has been coded as 1=male, 2-female. Recode it so that it equals 1=female, 0-male. Then, this variable's mean can be interpreted as a proportion. Therefore, if we were to poll the entire U.S. population, we would expect a population mean of around .5, or 50% female. Is there a significant difference between the gender representation in our 2018 GSS sample and our hypothetical U.S. population? Let's run a one-sample t-test to find out (SPSS does not differentiate z and t tests, so the t-test is used for both)