COMM 1 Lecture Notes - Lecture 12: Caffeine, Experiment, Internal Validity
Document Summary
Time-series designs track many observations over time, before and after a manipulation. What makes this different? better with more observations. Works for things better over a longer period of time, not every day think school testing, star, happens at regular intervals anyway. Find data before and after manipulation and observe changes. Why is this useful? improves the one-group pretest-posttest. Solves some threats to internal validity (testing, maturation) take the treatment away at a later point and measure again to be sure . E. g. , close the rec centers and see if crime goes back up. Have to be able to find similar things to study, two communities, two classrooms etc. E. g. , media literacy program to check, teach it to school a too and see if their scores go up and join b. Variation: give the comparison group the treatment for checking. Benefit of having two groups is that you can check. Non-directional (don"t know which way it is going)