We had noticed that the NPA test seemed to mirror real life, having the same ambiguities and problems of identifying the various types that one sees in actual individuals that we encounter. (The main problem is, of course, that the types often have one or two of the NPA traits in common.) This is not surprising, since the many real test results went into devising the test algorithm that makes the diagnosis. So, Test is a real, living and breathing individual, showing one of many faces depending only on how she answers the 50 questions of the test. Tweak the answers to the questions, and she shows a slightly different face. Exactly as in real life. Or, so she claims.
Test's premise was that since individuals with very severe MR would answer the questions in a random manner, then moderate MR could be simulated by tweaking the answers to the test questions with different degrees of randomness. The results of a fairly vigorous randomization (averaging each answer with a random number) indicated that some possible trends were:
1. toward non-perfectionism in NPA types having the P trait,
2. toward aggression in non-aggressive types, and
3. toward sanguinity in non-sanguine types.
4. little or no trend toward A− trait in dominant types having trait A (A, NA, PA, NPA).
The R score would decrease as the strength of randomness increased, so possible MR would be signaled by a low R score. But *caution*: a low R score does not necessarily imply increased randomness. There are many possible reasons for a low R score: see our previous posting:
Similarly, a high degree or randomization would give test results of NA−, but there are many real individuals who give test results of NA− with high R score.
« Back to index