Inconsistent mood congruent effects in lexical decision experiments.(Report)

During a lexical decision task, the mood congruent effect occurs when a participant responds more quickly and accurately to word items that are congruent with their current mood. The present study investigates the inconsistencies found for mood congruent effects in the lexical decision task literature. Previous studies that have successfully produced the mood congruent effect used a pseudohomophone nonword context while studies that have not successfully produced the effect used a scrambled nonword context. The purpose of this study was to assess if manipulating the nonword context would interact with the mood congruent effect in a similar manner as other semantic effects. A significant effect for nonword context was found. However, a significant mood congruent effect was not found.

**********

According to Bower (1981) memory consists of clusters of nodes that represent words, concepts, and events. Items in memory that are related to each other are connected. When a new event occurs, it is not stored in a secluded manner but forms connections with other related concepts and events. When an item in memory is accessed and becomes activated, energy travels along these connections and increase the activation of the related items. This increase in activation facilitates processing for these related items.

Bower (1981) proposed that memory contains emotion nodes that are part of the same associative network as words, concepts and events. These emotion nodes have connections to items that are related to specific moods. He suggested that two phenomena occur because of these connections. The first is known as the mood-congruent effect where people would tend to remember more about things that are related to the mood that was being experienced when the event occurred. The second is known as mood-state-dependent memory where people tend to recall more information about an event if their mood during recall is congruent with the mood that was experienced when the event occurred.

Bower's associative network model of memory has led to a number of investigations into how mood can influence lexical recognition. Based on Bower's model, it is assumed that when a person is in a particular mood, activation will spread to lexical items that are congruent with their mood. Therefore, responses to mood congruent items will be facilitated. Results from these experiments appear to be somewhat inconsistent. For example, Clark, Teasdale, Broadbent, & Martin (1983) performed a study where music was used to induce either a happy or sad mood prior to a lexical decision experiment. During the lexical decision experiment, the participant's task was to categorize strings of letters as being either words or nonwords (e.g. shoug). The word stimuli consisted of positive, neutral, or negative personality traits. Their results revealed no difference in response times for mood-congruent and mood-incongruent words and these response times did not differ from neutral word response times.

In a similar study, Challis and Krane (1988) performed a lexical decision experiment where participants were give positive and negative statements to read prior to a lexical decision experiment. Positive, negative or neutral trait adjectives were used during the lexical decision task. Their experiments produced a mood congruent effect for participants in the happy mood condition where happy adjectives were responded to more quickly than sad adjectives. However, they were not able to produce this effect for the sad mood induction condition.

Neidenthal and Setterlund (1994) also investigated the effects of mood on lexical access. Again, mood induction was performed using either happy or sad music prior to a lexical decision experiment. Words used during the lexical decision experiment included happy, positive, sad, negative, and neutral items. A mood congruent effect was found for both the happy and sad mood participants.