by Alan and Bo
Welcome to our new blog on correlation and causality. For four academic years (2003-04 to 2006-07 inclusive) we both taught research methods in the Department of Human Development and Family Studies (HDFS) at Texas Tech University. Bo has now moved to Penn State. Alan, who arrived at Texas Tech for the 1997-98 year, remains there.
Each of us has used a “validities” framework in teaching our methods courses (cf. Campbell & Stanley, 1963; Cook & Campbell, 1979), where internal validity deals with causal inference, external validity with generalizability, and construct validity with whether a test measures what it purports to measure (Linda Woolf has a page that concisely summarizes the different types of validity).
Bo one time said that, with reference to academics at least, internal validity is what gets him up in the morning and motivates him to come in to work. Alan’s passion is perhaps more evenly split between external validity (on which he operates another website) and internal validity.
On this blog, we seek to raise and discuss various issues pertaining to correlation and causality, much like we did during our frequent conversations at Texas Tech. In fields that study human behavior in “real world” settings, many potentially interesting phenomena are off-limits to the traditional experimental design that would permit causal inferences, for practical and ethical reasons.
Does the birth of a child increase or decrease couples’ marital/relationship satisfaction? Does growing up with an alcohol-abusing parent damage children’s development of social skills? How does experiencing a natural disaster affect residents’ mental and physical health?
For none of these questions could researchers legitimately assign individuals (or couples) at random to either receive or not receive the presumed causal stimulus. Much of our discussion, therefore, will be aimed at formulating ideas for how to make as strong a causal inference as possible, for a given research question.
By raising issues of how researchers might approach a given research question from the standpoint of internal validity, we hope to fulfill a “seeding” process, where our initial commentaries will be generative of further discussion and suggestions. We are thus permitting (and encouraging!) comments on this blog, for this purpose. We hope to learn as much (or more) from you, as you might learn from us.
In addition, we’ll write about stories in the news media that raise causal questions and review scholarly articles and books that do the same.
We recognize that issues of causality are implicated in a wide variety of academic disciplines. At the outset at least, we will probably stick closely to fields such as psychology, sociology, and HDFS. Later on, we hope to expand into other domains such as philosophy and legal studies (within the law, many states have homicide or wrongful death statutes with wordings that allude to situations in which someone “causes the death” of another).
We invite you to visit this blog often and chime in with comments when the feeling strikes. Requests from readers to write lead essays as guest contributors will be considered (or we may even invite some of you to do so).
Thanks for stopping by!
1 comment:
This blog is a great idea. Although I have taught research methods from a validities perspective using most recently Shadish, Cook, & Campbell's book on experimental and quasi-experimental designs, I have had difficulty identifying a source for teaching validity of correlational designs. The references you cite on the blog will be helpful.
Post a Comment