Therefore, non-simplicity need not be listed among the theorem's conditions, though it is needed in the variants dropping path-connectedness.
We gradually dropped paths if they were nonsignificant until only significant paths remained.
After all the insignificant paths in the hypothesized model had been dropped, paths were added at each step to improve the goodness-of-fit statistics based on the modification index and pre-set assumptions.
The second rule was to drop adjacent paths (e.g., between reading differences at ages 10 and 12) before longer term paths (e.g., between reading differences at ages 10 and 16).
If dropping the path resulted in a significant change in model fit, the path was deemed significant and retained for the final model.
First, we dropped the path sets in the reverse order to that described earlier, beginning with the contemporaneous covariances and ending with the within-trait reading paths.
For within-trait reading difference associations, we could drop four paths without significant loss of fit (all resulting changes in fit are shown in Table 2).
And if the paper could be etched with tiny channels so that the drop followed a path, and if that path were mined with dried proteins and chemically triggered dyes, the thumbnail-size square could be a mini-laboratory — one that could be run off by the thousands on a Xerox machine.
A final variant drops both path-connectedness and pair-negatability while imposing both systematicity and monotonicity (ibid).
The first rule was to begin by dropping these paths (e.g., between reading differences at ages 12 and 16).
