The principle of designer tests is that of using a subject's own semantics
rather than lists of words that may or may not be relevant or even understa
ndable for the subject. The Personalised Emotional Index (PEI) is a prototy
pe designer test, in this case a mood test, that uses words that the subjec
t chooses from a list of suggestions within mood categories. Each person's
test is custom made from familiar and understandable words from his/her own
vocabulary. Such a test has much face validity, can be succinct and has co
mprehensibility for the subject. The results obtained when using this test
at the same time as the Profile of Mood States Bipolar Version (POMS-BI) we
re very similar (e.g. in a regression analysis, the 'elated-depressed' vari
able predicted present overall mood on both tests (POMS: t = 5.25, p < 0.00
0, PEI: t = 5.84, p < 0.000) with a high correlation for total scores (r =
0.82, p < 0.000). The PEI results were correlated within the two week inter
val (r's about -0.74; p < 0.000) and reasonably but not highly correlated o
n retesting some months after the first testing (r's about -0.25; p < 0.000
). It was successfully used to differentiate mood variables from a group co
nsisting of caregivers of people with schizophrenia (n = 30, producing 399
days of data) and a group of unselected controls (n = 62, producing 1080 da
ys of data). The lest appears to have validity, reliability, comparability,
and utility. (C) 1999 Elsevier Science Ltd. All rights reserved.