A brief word, I promise, about psychological statistics.
Every answer given by a participant in my study had a number value. These values, instead of actual words, were put into a statistics computer program called SPSS. The computer takes complicated mathematical formulas, and uses them to find relationships among all of the measures and their values. It's really cool. It would take days for me to do what the computer program did in just seconds.
For you statistics fans out there, SPSS computed a Pearson r value for each correlation of the CHAOS measure with the other measures.
The scatterplots were so exciting! I could clearly see that the CHAOS measure was negatively correlated with four of the measures, and positively correlated with one measure. As chaos scores went up, scores on 4 measures went down, and scores on one measure went up, as predicted. But...what was even more exciting than the scatterplots, which only compared the scores for the sample of 52 people, was the test of significance. Without that, I could only assume that a correlation existed for those few people, but what I really wanted to know was, would those correlations exist in a larger population?
When the computer did a test of significance, the chances that the results were coincidence, ranged from less than 1/1000 to only 7/1000. That means that the study's results would be found in a larger population. That's significance, and what I was ultimately looking for.
For all of the statistics fans out there, here is a table of the important values.
Coherence Communication Adaptation Routines Emotion & Criticism
FOC FPSC FAS FTRI FEICS
Pearson Correlation -.518 -.646 -.387 -.401 .371
Significance (2-tailed) <.001 <.001 .005 .003 .007
High scores on both the CHAOS AND the FEICS indicate a dysfunctional situation. High scores on FOC, FPSC, FAS, AND FTRI indicate things are going good at home.
CHAOS negatively correlated with the first 4 measures, as predicted & positively correlated with the last measure (FEICS), as predicted.
One more interesting statistic. If you take those Pearson values and square them, you get a % measure of convergent validity. That means that the CHAOS measure is measuring some of the same items that the other measures are measuring, but only some. That's a good thing! For example: The FOC r value = -.518. Rounded & squared it = 25%. That means the CHAOS overlaps with the FOC (a measure of organization) by 25%. That's fabulous!