People who experience delusions have been found to request less information prior to making a decision than control participants on tasks that are unrelated to the theme of the delusion (Huq, Garety & Hemsley, 1988). Two studies investigated whether people with delusions have an absolute deficit in reasoning or a more specific data-gathering bias. In Expt 1, 12 people with delusions, 12 people with depression and 12 normal controls were shown the results of spinning a supposedly biased coin. The evidence provided varied in the number of heads to tails. In normal controls a high ratio of head to tails produces a high estimate that the coin is biased. In this experiment, where the evidence gathered was predetermined by the experimenter, all groups of participants were shown to reason in a similar way. Experiment 2 tested whether a difference would exist between the groups in conditions where participants were free to determine the amount of evidence seen, in contrast to when all of them viewed the same evidence. Two jars of beads in opposite but equal ratios (e.g. 85:15, 15:85) were shown to 15 people with delusions, 15 with depression and 15 normal controls. On the basis of beads being drawn one at a time, it was the participants' task to determine which jar they came from. When free to decide when they wished, people with delusions decided on the basis of less evidence than the other groups. However, as in Expt 1, the group with delusions did not differ when made to view the same amount of beads as other participants. Therefore, people with delusions have a data-gathering bias rather than a difficulty in employing the data in reasoning. This "jump to conclusions' bias generalized to a less discriminable ratio of beads (60:40), and was not a consequence of impulsiveness or memory deficit.

People with delusions have been shown to have both generalized (Huq, Garety & Hemsley, 1988) and content-specific biases in reasoning (Bentall, 1994). Our concern here was whether the hastiness that has been found when people with delusions reason on relatively abstract tasks would be present on a more realistic task. A second concern was whether reasoning with salient or emotional material would increase the hastiness bias in people with delusions. Two versions of a probabilistic reasoning task were used to study the data gathering of people with delusions. The first version employed realistic but emotionally neutral material. People with delusions requested less evidence before making a decision than psychiatric and normal comparison groups. Therefore, the hastiness found previously with abstract materials was seen to generalize to a more realistic task. In the second version participants were required to reason with material that had an emotional content and may have been regarded as being personally meaningful. In this condition all groups reduced the amount of evidence requested before making a decision.

The norms which follow are a collection of free association norms to four categories of emotional words (happiness, sadness, anxiety, and anger) and a corresponding set of non-emotional words, matched for frequency of occurrence and for word length. The norms are accompanied by ratings of emotional content, and presented in decreasing order of emotionality. They were collected from a total of 300 Reading University undergraduates over a period of three weeks in 1985.

Background: Previous research has indicated that persecutory delusions and depression may share similar cognitive biases at implicit levels of processing, but differentiate at explicit levels, supporting the theory that paranoia may have a protective function against underlying negative schemata. The study aimed to investigate attentional bias and both implicit and explicit memory biases for personally salient and standardised emotional stimuli in persecutory delusions and depression. Sampling: 36 participants, with 12 in each group, were interviewed in order to generate personally salient stimuli to be employed within the cognitive tests. Standardised emotional stimuli were additionally employed as a control. Participants completed two probe detection tasks, one including personally salient stimuli and one including standard emotional stimuli. Memory for the stimuli presented in this task was assessed by a free recall task (explicit memory) followed by a word completion task (implicit memory). Results: On an implicit memory task, both the deluded and depressed groups displayed comparable retrieval of positive and negative words. However, on the explicit memory task, the depressed group demonstrated a bias for negative stimuli, whereas the deluded group demonstrated a bias for positive stimuli. The groups did not demonstrate an attentional bias for personally salient information. However, an attentional bias for standardised emotional stimuli was found in the depressed group, although this was not specific to either negative or positive stimuli. Conclusion: The results indicate that depression and persecutory delusions may share similar patterns of processing at an implicit level but differentiate at the explicit level, which may be indicative of cognitive avoidance of threatening stimuli in psychosis. However, this does not seem to be a feature of automatic attentional processes in people with persecutory delusions. Implications for further research are discussed.

No abstract