Chapter 5 Alternatives to Experimental Research in Social Psych Profound events are worthy of serious empirical scrutiny even if they cannot be examined with the meticulous research designs associated with laboratorybased research in social psychology (i.e. natural disasters, terrorist attacks). Pennebaker and Harber: collective coping based on selfreported thinking and talking following the earthquake: emergency phase, inhibition phase, adaptation phase. Thoughts about the event remain relatively high during first two phases but dissipate in the third phases, where people act as if the event is over. In contrast, actual talking about the event declines from the end of the emergency phase through the adaptation phase. In the inhibition phase, lack of talking about the event increased thoughts about it, mentally stewing over the events. Problems of leaving the lab: No random assignment to conditions No controlled manipulation of an IV No conditions representing distinct levels of an IV No sensitive DVs that allow researchers to link cause and effect No control group(s) or limited use of comparison groups No opportunity to debrief research participants. Broadly, research in the real world lacks the rigor associated with the lab where between or withingroup differences caused by manipulating an IV can be cleanly and clearly documented. The advantages of nonexperimental research include: Thinking broadly and creatively about social behaviour in actual social contexts. Testing out ideas uncovered in the lab out in the field (and vice versa). Developing new and creative DVs and ways to measure behaviour. Having opportunity to examine how people respond tocope with a real social or natural event. Using social psychological theory to explainimprove some practical issue in the social world. Observational research: watching people engage in normal behaviour in public settings. Does not interfere in the situation observations are made unobtrusively so that individual(s) being observed do not alter their actions, but some record is being maintained. Largest advantage is that it can be done virtually in any public place, can be used to generate ideas. Largest disadvantage is that it does not allow researchers to determine causality. Correlational research a measure of association between two variables (and the degree to which that correlation exists). Positive correlation: as one variable increasesdecreases, the other does the same. Negative association: as one variable increases, the value of the other decreases. Zerono association: no pattern of covariation between the variables being considered. Measured by the correlation coefficient (r): 1.0 to +1.0. Passive design measuring how variables covary without any direct intervention or manipulation. Can be helpful for determining the likely predictor of a measure correlated with a personality trait. Disadvantage: correlation does not indicate causation the presence of an association does not tell us anything about the causeorigin of the relationship. Quasiexperimental research designs: close to, but not quite an experiment because it lacks one of the following experimental control, random assignment to condition, lack of a representative control or comparison group, and manipulation of an IV. These are not always possible when conducting research on groups (cancer research: good treatment vs. bad treatment not ethical).