PSYCH 100A Lecture 10: Psych 100A W5L1 2/7/17
Document Summary
We create two hypotheses, tested against one another. We test the null hypothesis trying to disprove it. Use a certain level of uncertainty and potential errors. Slightly different formula: we use standard error instead of standard deviation and we"re no longer comparing individual scores, we"re comparing the sample mean. Iq example: average iq = 100 with sigma = 15. Z counts the number of standard errors any sample mean is above the mean specified by the null hypothesis. If zobs is greater than critical z, we reject the null hypothesis. We use the sampling distribution of means to compare the z values. The sampling distribution of means is narrower than the sample distribution. In this case, we use a z-test not a t-test because the population standard deviation is known. Upper limit for iq = mean + (critical z)(standard error) Critical z = 1. 96 for when alpha = 0. 05 (95% confidence level)