PADP 7110 Lecture Notes - Lecture 8: Internal Validity, Focus Group, Time Series
Document Summary
Lecture 8 : research design non-experimental: attributes, researcher does not have control over events, tests whether one variable relates to another in a way that suggests causality, generally . Don"t have multiple observations over time: includes qualitative approaches (e. g. case study, focus group, does not control for threats to internal validity. Can collect other valuable, detailed information: single-group, pretest posttest design. Can gather data to answer specific questions about a program. Can combine knowledge of the program with other information to make a plausible case that it is effective. Identify questions to investigate further: nonequivalent groups posttest only design. Can"t tell if differences between groups existed prior to the intervention or if any changes took place in each group. Can potentially identify places of program that are working and parts that need improvement. 2: one-shot survey, qualitative approaches > designs for description, cross-sectional designs (surveys, forms, database) Provide a snap shot of population of interest at one point in time.