Reasoning and Affective State in Terms of Psychology

Affective state is a hypernym encompassing emotion or mood; emotions and moods are a set of inner psychological ‘feelings’ that are instinctive and subjective. Emotions are intense and brief whereas moods are longer lasting and stable however less intense, thus neither of these states can be directly observed. This makes them extremely difficult to measure in a psychological discipline, leading to the study of affective states being dismissed in early psychology, particularly in the fields of cognition and behaviourism as they were seen to complicate research (Gardner 1985). However more current research relating to eye-witness testimony and flashbulb memories, suggest affective states and cognition are intertwined and to perceive them in isolation would be problematic Therefore, it is of benefit for researchers to look at whether affective states hinder or facilitate cognition/reasoning due to their real-life consequences. Affective states can be difficult to measure as manipulating and assessing them is challenging as individual differences may come into play, and researchers also use different measures to manipulate and assess therefore making it hard to draw valid conclusions between any two studies. Though there is an important distinction between the way these affective states are manipulated, which can either be integral or incidental studies; incidental is when there is no link between the affective state and the reasoning task, whereas in integral studies, the affective state is linked to the content of the reasoning task thus serving more of an interaction. In addition, researchers can manipulate whether it is the affective state, or the content being reasoned that is inducing the state which can vary from study to study. Research into reasoning can also measure more ‘broad’ affective states, e.g. positive versus negative or more specific, such as anger versus surprise.

Reasoning is the practice of consciously drawing conclusions on current information via logical thinking and adjusting current conclusions if there is novel information available, for a plausible outcome. However, if reasoning is universally rational it should be resistant to the influence of affective states; in order to test this, deductive reasoning is commonly used (Blanchette & Richards 2010) in which inferences are drawn from a set of syllogisms in order to assess validity.

Research has been performed to assess various types of emotion on reasoning abilities and whether one emotion provides a marked response (Blanchette & Leese 2011). Initial research simply suggests that people reduce their accuracy when reasoning if they are drawing conclusions on emotional content or in an emotional state, in comparison to neutral content or in a neutral state; this is supported by Lefford (1946) whom depicted emotion as unfavourable to reasoning accuracy (Lefford 1946). In early research conducted by Lefford, participants reasoned for emotional and unemotional syllogisms and found that when reasoning about emotional content, accuracy was reduced, suggesting that when a syllogism causes an affective arousal, this impedes reasoning accuracy in some way.

Studies looking at the more rudimentary positive and negative mood relationship with reasoning, have been conducted by Oaksford et al (1996). They created an incidental study manipulating participants’ mood before reasoning, using film clips. They found provoking both positive and negative mood suppressed accuracy when completing a Wason selection task in comparison to those in a neutral mood. Although this suggests both positive and negative mood have an effect on reasoning, it may be the case that a specific mood leads to that specific response not just the overall ‘positive’ or ‘negative’ connotation; a positive mood can encompass happiness or surprise, yet they represent completely different moods. Therefore, this study lacks the in-depth evaluation of specific moods as it provides a broader basic overview.

Studies for specific manipulated moods can therefore provide greater accuracy. Blanchette and Richards (2004) conducted an experiment testing the emotional content of words when reasoning; they used either emotional words (representing anxiety, sadness and happiness) or neutral words in place of P and Q syllogisms. They found that participants were more accurate in reasoning neutral words, rather than emotional, showing that emotional content can affect reasoning also, not just emotional state. In their second experiment they used classical conditioning in order to prime participants with neutral or non-words for emotion. They undertook this by pairing the non-words with pictures that were perceived as positive, negative or neutral in their emotional content. They used these non-words as they believed it would be easier to prime participants with unfamiliar words. They then used these primed words in replacement for the P and Q syllogisms, showing the same effect as Oaksford (1996), that participants were more accurate when reasoning with neutral words in comparison to the primed emotional words. This study therefore concluded that people were sensitive to the emotional content when reasoning by showing a diminished response.

Together, this research by Oaksford (1996), Blanchette and Richards (2004), shows emotional mood and emotional content incriminate reasoning accuracy supporting earlier research by Leford (1946). Though an obvious issue with this research, along with other laboratory studies, is that they lack ecological validity when manipulating affective states. In addition, Blanchette & Richard’s (2004) study uses only word content but emotion in real life is far more complex to express than the use of single words, which may not have provided the desired effect. In addition, when mood is manipulated, it may become a comorbidity with another mood and may not necessarily erase the participant’s previous mood/s. This can be an issue in methodology, as it may be the case that another mood, besides the one being manipulated, is influencing the reasoning accuracy.

Due to the above issues with laboratory studies, real-life affective states are imperative to study, firstly these studies are preferred as there is no manipulation needed to prime the participants in their affective state, resulting in more authentic results. When manipulating in a laboratory, participants may be primed to feel a certain affect eg.by seeing a sad film clip, yet when later tested with their reasoning ability, the contents they are to reason, may not be related to how they were primed initially thereby parting their affective state from the semantic content (incidental); this issue can be more easily avoided when using real-life affective states and integral studies (e.g., Bodenhausen et al., 2000; Oaks ford et al., 1996; Palfai & Salovey, 1993).

Integral emotion studies show more of a direct effect, testing the link between emotional state and emotional content. Blanchette and Richards (2007) looked at real-life terrorism incidents as they are alarming to the individual and therefore there is no doubt they will induce deep emotion. The 7/7 London Terror Attack bombings were used as the basis of real-life emotion in this study. Participants in Manchester (UK), London (UK) and London (Canada) reasoned deductively with syllogisms that held either neutral, generally emotional or terror related content. Participants were tested one week after the incident and then six months later. All participants self-reported the measure of their emotion. Findings showed that London (UK) participants reported higher on their self-reported emotion levels than the rest of the participants from Manchester (UK) and London (Canada). Further results however did not show that emotion caused impaired logic in the London (UK) participants as they were specifically more accurate when reasoning about terror syllogisms; this was then followed by Manchester then London (Canada). The closer the participants were to the 7/7 attack the higher their self-reported emotion; London (UK) participants were higher in fear and in contrast also higher in positive emotions, showing that differing emotions can exist along-side each other. Overall neutral problems were the most accurately reasoned within all groups in comparison to emotional, however London (UK) participants were just as accurate when reasoning terror related content thus concluding that higher reports in emotion, either negative or positive does not reduce reasoning ability as shown with the previous studies. As this study was integral, this suggests that when reasoning syllogisms direct to the crux of the individual's emotion, enables them to reason with even more accuracy. However, it is worth noting that terrorism can be subject to stereotyping towards a certain group, which may lead to heuristic processing thus illogical conclusions when reasoning, which might indicate a flaw in the methodology of the study and the individual’s working memory.

In a similar study Blanchette & Campbell (2012) also looked at the effect of real-life emotion and its impact on reasoning, using ex- war veterans. They measured ex-veterans reasoning on neutral, emotional and combat related syllogisms. They too, found a similar pattern to Blanchette and Richards (2007) study that when the reasoning is related to the participants specific emotion/experience, they reason with better accuracy compared to neutral content, in this case combat experiences. This therefore shows an integral response, that dependent on the emotion, you get a facilitated effect when there is an affiliation between the affective state and the reasoning content. However, those participants with PTSD were less accurate for their reasoning in all conditions compared to the other groups, but still had an advantage for combat related content. This is no surprise as Ex-service men have higher rates of PTSD which serves a correlation with working memory ability which may implicate reasoning accuracy which can be a non-pathological consequence PTSD which severe emotion can cause. They also found a link between those with more intense experiences and found they had reduced accuracy for emotional content and combat related experience which suggests they may have become more ‘immune’ to the effects of emotion having to deal with it on a higher basis, hence why they reasoned with it the same as neutral content. This suggests when individuals have extreme emotion this may taint their overall performance in reasoning, suggesting that when emotion crosses a certain caveat it can have negative effects no longer benefitting reasoning. The findings above may by accounted for by relevance, as integral emotion is deemed more relevant than incidental emotion. When the emotion and reasoning content is integral, more attention may be paid which increases accuracy compared to more irreverent integral reasoning, deferring attention though, this was not seen in those with intense experiences. Therefore, dependent on the emotional context/content, will result in how much attention is given to reasoning conclusions. Nonetheless these findings may have been due to familiarity as those whom have had more combat-experience will therefore perform better when reasoning in their field, though emotional/neutral reasoning should also have seen this facilitation as they are more ‘everyday emotions’, though they did not.

Therefore, when the emotion is linked to the content of reasoning there seems to be a benefit to reasoning accuracy. However, research by Jung, Wranke, Hamburger & Knauff (2014) contradicts these findings. In their 2014 study, participants completed logical inference problems. They tested participants with exam anxiety or those with a spider phobia; these participants solved problems linked to their affect. Those with a spider phobia resulted reduced accuracy when reasoning with spider content compared to neutral. Those who were exam anxious resulted in the same performance accuracy to those non-anxious therefore showing a neutral effect of emotion. The findings therefore conclude that no matter if there is high relevance to the participants emotion, there was no facilitated effect present for reasoning, contradicting the findings shown by Blanchette et al above. However, phobias are extreme irrational fears which may not be equivalent to everyday emotions and are therefore somewhat an anomaly explaining why this research does not represent the previous findings. In addition, a phobic’s goal is to avoid their fear which may trigger their phobia explaining the reduced accuracy (DSM-5 2013). The study testing exam anxiety consisted entirely of students and although some reported no exam anxiety at all, this is unlikely to be the case as they might have self-reported inaccurately. This might show that perhaps the non-anxious performed equally to the anxious as they had similar anxiety levels thus showing the neutral findings. In addition anxiety may use maximum capacity in working memory, therefore resulting in decreased accuracy.

In more deep-rooted emotion, Channon and Baker (1994) looked at reasoning for syllogisms incidentally in depressed participants and found that these participants reasoned with considerably lower accuracy in comparison to controls, even when the reasoning was not emotional, there is still an influence of the individuals internal emotion on their reasoning success, this may suggest WM and cognitive function influence on reasoning ability, as those whom are depressed are lower in this. Therefore, suggesting other explanations of individual differences in reasoning accuracy, that some individuals may have reduced central processing and working memory ability.

Explanations for why affective states may have effect on reasoning is not a direct effect per se, but due to the pressure these states place on working memory, as the amount of capacity these states take up, leave less capacity for reasoning, which is reliant on working memory. This means that affective states work in the same way as if you were to complete two tasks at the same time resulting in an increasing demand on WM.

An explanation for all these contradicting findings above, can be explained simply due to the differing methodologies. Affective states, emotion and mood are hypernyms and are therefore wide-ranging in that sense. Specific moods however can have very different effects on individuals that you cannot measure introspectively. Therefore, these specific affective states can vary emotion to emotion and intensity to intensity thus effecting reasoning accuracy as although all participants may be experiencing the same emotion/mood they may not be experiencing it in the same intensity, showing different results. However, if there was an impact of affective states on reasoning, it should not matter how small the emotion as the lowest baseline should show an effect. Therefore, for more accurate insights, specific affective states should be analysed with tighter control and not just measured in the broader sense.

In a more complicated explanation, reasoning is not a straight-forward process as it recruits many cognitive resources in order to complete, such as working memory, attention span and even more complex processing such as constituent processes, priming effects, computational capacity, reflective processes and deliberative processing. Affective states and their role in these individual central processes needs to be conducted in order to see what they are precisely provoking. To conclude, it is not as simple as affective states acting on reasoning, yet all the other resources that combine to make up reasoning, with each affective state having a different effect on each resource, as well as the intensity (Blanchette & Richard 2010). There is undoubtably a complex interaction of affective states throughout the reasoning process, and some studies may have been able to disentangle the relationship better than others which may explain the contradictory findings.