35
Dealing with Dynamic Systems:
Figure 3. Causal structure of the ALTθL simulation. The signs in parentheses are valid only
in the “mismatching version” (see text). The Roman figures indicate the numbers of the
partial systems which are mutually independent.
Independent and dependent variables. Experimental variations were made
with respect to the following factors. (1) Knowledge compatibility: Two
versions of ALTOL were realized such that the signs of the parameters of
two of the four exogenous effects were either in concordance with the
assumed prior knowledge (matching condition) or disagreed with it (mis-
matching condition). The question was how the semantic embedding inhibits
or facilitates the identification of two systems with identical structures. (2)
In one condition subjects had to control all four endogenous variables on the
last block, whereas a second condition required the control of only two of the
variables. (3) The experiment also investigated whether control of the system
would be influenced by the number of exogenous variables by which subjects
could manipulate the endogenous variables. This was a within-subjects fac-
tor. The ALTOL system consisted of three independent parts (marked by I,
II and III in Figure 3) with a 1:2, 1:1 and 2:1 relation between exogenous and
endogenous variables. (4) The last factor manipulated the user interface of
the system: In one condition the variables were presented numerically (Exp.
4), the second condition used a graphical presentation which contained the
same information (Exp. 5; see Fig. 4).