SOCECOL 13 Lecture Notes - Lecture 13: Sampling Frame, Dependent And Independent Variables, Test Statistic

• School: UC-Irvine
• Department: Social Ecology
• Course: SOCECOL 13
• Professor: Kenneth S.Chew

Document Summary

Short answer format write on test form only. One 4 x 6 card write or print anything (both sides); may use glue but no extensions, foldouts, or other tricks to increase card area. Bias polls are always at least a bit off. Please think of a measure or question that is valid but not reliable. Please think of a measure or question that is reliable but not valid. <> statistical significance. An observed relationship in a sample is statistically significant to the extent that it is unlikely to have occurred unless there really is a relationship in the population: statistical significance is not equal to real world importance. When it is evenly distributed, no difference across rows = small x^2. What we would expect if nothing is happening; if the treatment isn"t working = even distribution. X^2 = upper right hand expected with upper right hand observed distribution (made up of cell-to-cell differences)