COMM 101 Lecture Notes - Lecture 23: Face Validity

• School: U of M
• Department: Communication Studies
• Course: COMM 101
• Professor: Julia Lippman

Document Summary

Sources of error in design are not ignorable - the impacts of error: seemingly contradictory results, seemingly inconsistent results, explanations about unusual groups, answers to different questions than those posed. What are the assumptions: about the measures, sampling, procedures, data, statistical tests , these are often wrong: Questions may not assess what the researcher thinks they will. Samples may be representative of the wrong populations. Researchers may assume that the data is clean when it is not. Take away #2: know the assumptions and discuss them. Successful replication: the goal is to build the strongest possible evidence for a hypothesis, if a hypothesis is true, then evidence for the hypothesis should be reliable. Two key types of replication: exact replication - reliability of study. Same procedures and operationalizations: conceptual replication - validity of study. It"s easy to lie in the conclusions because scienti c results are subject to interpretations. There also may be many explanations for what is observed.