“Estimating the Magnitude of Experimental Effects”, 1969 ():
[Discusses the increasing awareness that the mere statistical-significance of an experimental effect is insufficient to warrant the conclusion that the effect is large and practically important. ]
…Yet more complicated designs are frequently employed in psychological research: The treatments may assume a factorial structure; some random factors may be nested within the levels of other factors; etc. No simple formulas for estimating the magnitude of experimental effects in all such situations seem possible. Rather, the investigator must examine the column of expected mean squares in his analysis of variance table, identify the bona fide components of the variance of an observation, and use as an estimate of total variance that linear combination of mean squares appropriate to his design.