Day 1
Breakout room discussions: Reversing cause and effect, inver:ng correla:ons, thinking about mutual causa:on, third variables.One very produc:ve cri:cal thinking tool is inver:ng the way we typically think about a cause and effect. This is can some:mes reveal assump:ons we have made, and provide new ideas for tes:ng.Take a causal claim rela:ng two variables and try to do the following:
1)Explain how the first variable may cause the second (e.g. Frequent exercise causes elevated mood (perhaps because of the release of endorphins)
2)Explain how the second variable may cause the first (e.g. Elevated mood causes frequent exercise (perhaps because when people feel good they enjoy exercise more)
3)Explain how they might ‘cause’ each other in some dynamic rela:onship. (e.g. Elevated mood increases exercise, and that in turn tends to improve mood, and perhaps beQer health, and self-esteem)Another key tool is genera:ng alterna:ve explana:ons for a rela:onship
4)Iden:fy a third factor that might explain the rela:onship observed (e.g. Certain people are gene:cally predisposed to elevated mood and propensity to exercise frequently, or, people are both happier in the summer and exercise more frequently, and the opposite in the winter)Finally, iden:fying addi:onal factors that might influence an outcome variable, yet not be the sole cause is another task for cri:cal examina:on of a situa:on. Not all factors that have an influence on one of your variable are confounds, or true alterna:ve explana:ons, they merely explain a liQle bit more about an outcome.
5)Name one factor that influences the variable you want to call the ‘effect’ (rela:onship status influences mood, :me of day influences mood, drug use influences mood)
Day 2
Breakout room discussions:Understanding causal rela:onships takes understanding the sequence of events in :me, and the ability to rule out alterna:ve explana:ons
Discuss how Longitudinal designs help us move towards understanding the temporal precedence
Discuss how mul:ple regression can give us some Sta:s:cal Control over other factors.
