Ratings are presented for 660 stimuli from word-association lists on each of five scales: goodbad, pleasant-unpleasant, emotional-neutral, concrehabstract, and easy to associate todifficult to associate to. The ratings are shown to be highly reliable, and to agree well with previously collected norms of a similar character. The intercorrelations of the five scales with one another and with word frequency are reported.As research on the determinants of word-association performance proceeds, more and more stimulus characteristics become implicated, and the need for normative data increases. Future experiments will benefit from the availability of words with known positions on several important parameters, since it is often important to control for one factor while manipulating another. At the same time, the utility of previous findings is limited by the absence of normative data on the words used as stimuli by earlier researchers. In order to meet at least part of this lack, the words to be rated in the present study were chosen because they had served as stimuli in previous word-association experiments.Ratings on five seven-point scales were collected. Four of these scales have been used in previous normative studies. These are the scales for goodness (G), pleasantness (P), emotionality (E), and concreteness (C). The Gfth is a scale of associative difficulty (AD), defined by the polar terms 'easy to associate to' and 'difficult to associate to'. There is no exact precedent for a scale so defined, but it clearly has affinities to ratings of association value and meaningfulness (Cieutat, 1963;Noble, 1961). METHODStimulus words. I n all, 650 words were rated. A few unintentional omissions apart, they represent all stimuli used by Kent Brown (unpublished). Preparation of booklets.The ratings were made in booklets consisting of a two-page introduction explaining the rating procedure, followed by 15 pages of rating scales. Seven-point rating scales were used, 25 to a page. The stimuli were printed on the left of the rating scale, and the opposed terms defining the scale were printed at the top of the page, aligned with the extreme points of the scale; in half the booklets the scale was reversed. Since the total number of words was judged to be too great for subjects to be asked to rate them all, they were divided into two sets of 325 words. Each booklet contained one of these sets. Each word occupied a standard position on its page, but the order of pages was randomized. In addition, two further pages were included to provide data on the reliability of ratings. Of the 50 words on these two pages, 25 came from each set of 325 words. Thus each subject rated 25 words from one set a second time, and 25 words from the other set once. This enabled us to estimate retest reliability of the ratings both with the same raters and with different raters.Subjects and test setting. The numbor of subjects rating each word on any scale varied from

In a survey of 140 patients in the gynaecology wards of a large general hospital, the subjects were asked whether they suffered from any of the common hay-fever type allergies, for example, to flowers, bedding dust, etc. An incidence of 20 per cent was established. Surgical and biopsy reports on this ward population established an incidence of malignant conditions at 28 per cent. These two groups of comparable age turned out to be mutually exclusive.Reports of pregnancy nausea were also taken. Previous susceptibility to this varied from none to severe, and correlated positively with reports of common allergic type reactions, and consequently, inversely, with liability to gynaecological cancers in later life. The possible significance of this result in the studies of cancer is discussed.

No abstract