Three research dezigns are presented and the infiuenoss of
these various threzta on ezch are discussed. (see Tuble 1, p, 6)
»Tho "lnesShot" Case Study

x 0

The authors denigrate this as the lezst worthuhile desim.
The laek of oontrol and lack of comparison exespt with "inplieit
ocamon knowledge" lead to the "error of misplaced precision."
This design also suffers from all of ihe threate to vžidity exeept
those that involve multiple measuremonts.
»Tho dne-Group Pretest-Posttest Design

0 X 0

While this design is deemed betto: than doing nothing, the
authors use it as an exprple of nany of ibe thrests to internal
velidity, Of all of the thrsate, only sslection and mortality are -— .
controlled. As an interesting note, it is the feature of "experimental
isolation" that allows the physicol se!encos to often gain control over
the threat of history uhen using this design,
»The Statlo.Gronp Comperison

s.a
Here, one group experlencvs the treatnent, snd the other does
not. The dotted iine shows tha!, no means are used to test whsther the

are
differences betumen groups /Žz čue to treatment. This design centrols
for all threats, except for sr leotion-and mortality.