Menu Table of Contents Previous

IV. Meaningfulness Maximization

It is not enough simply to maximize the likelihood that someone else will get your results if they repeat your procedures; or that they will get your results if the procedures vary in some way; or that you made every effort to find small effects. Your findings must also be meaningful. If you are an emergency room medic it is more meaningful to discover arterial bleeding than ingrown toe nails. If you are charged with finding the sources of noise in an apartment it is more meaningful to discover a free running jackhammer than a dripping faucet.

This basic concept takes two major forms in psychological research: 1) it is important that variations in your independent or predictor variable account for a large proportion of the variation in your dependent or predicted variable and, 2) You must maximize the ability of your paradigm to make sense of all the "booming and crashing" apparent randomness in nature. You can maximize the meaningfulness of your research by maximizing your understanding of the existing paradigm and/or making a better paradigm..

Maximal reliability, generality, and detectability are necessary for maximal meaningfulness, other things being equal.

Removing noise from a room illustration.

Regression line illustration.

Venn diagram illustration.

Partitioning history test performance illustration.

This, of course, requires that the effect is still apparent or an apparent effect is obtained.

Start with few animals. Variety of subjects, apparatus and procedures.

The better you know the existing paradigm, the better you will be able to use the knowledge base to solve problems.

There are foundation assumptions underlying every paradigm. You cannot understand the paradigm without them.

Paradigms provide machinery with which events in nature can be understood. They are a core aspect of a paradigm.

How can we assure a long-term impact for our research?

This is normal science.

This is revolutionary science.



Menu Table of Contents

Send comments/criticisms/speculations to

Date Last Reviewed: November 17, 2002