Abstract:
A mixed factorial design was used to test the hypotheses; a) participants with high caffeine levels perform better than those with low caffeine levels, b) Participants with high positive affect (PA) scores are more likely to perform better than those with low PA scores and c) high caffeine intake and high PA scores will show significant interaction effects on the stroop bias score. 243 Coventry University students split into two groups based on their PANAS score. The low PA score group (12-21) contained 159 participant with a mean age of 20.31 years. The high PA score group (22-36) contained 84 participants with a mean age of 19.83 years. Participants completed a shortened version of the PANAS questionnaire, ingested a tablet containing either 200mg or 400mg of caffeine, after 20 minutes they completed a stroop test. The results revealed no main effect of caffeine on stroop bias scores, PA score on stroop bias scores and no significant interaction effect of caffeine and PA on stroop bias scores.
Introduction:
Caffeine is one of the most commonly used psychoactive substances today (Snyder & Pamela, 1983) available to all ages (Jacobson & Kulling, 1989) through various products including Coffee, Pepsi, and pain relief medication. Tea and coffee being the most popular. Many believe small doses of caffeine such as a cup of coffee which contains approximately 100 mg of caffeine enhances psychomotor performance, while higher doses may cause insomnia, anxiety and irritability (Jacobson et al, 1989). Severe doses can cause nausea diarrhoea and even death. Given its accessibility and rate of consumption, the effects of caffeine on cognition are important to understand.
Edwards et al, (1996) designed an experiment to determine the effects of caffeine on cognitive performance using a Stroop task. Results showed no significant effect on performances in the colour-word version nor the numerical version when stimuli were presented on either computer or card version. However, they observed significant practice effects in the within-participants design when using the card version. Differences were also found within the between-participants design in both the card and computerised versions of the Stroop. Researchers concluded 125mg and 250mg doses had no significant effect on Stroop performance. Practice effects potentially proved an issue when measuring effects of caffeine on cognitive performance. The following study improves this method in several ways, a higher dose of caffeine will be used to attempt to see potential effects more clearly and the research will try to counteract practice effects using practice trials of the stroop, finally, participant’s mood will be recorded to see if mood affects performance.
Positive affect (PA) is the extent to which individuals experience positive feelings or emotions such as happiness. Previous research into the effect of PA on cognitive control produced mixed results. Van der Stigchel, Imants, &, Ridderinkhof (2011) studied whether PA increased the ability to suppress a reflexive saccade in an antisaccade task. Results showed participants made fewer erroneous pro saccades in the PA condition compared to the control condition. Further research found PA increases stroop interference due to heightened distractibility (Rowe, Hirsh & Anderson, 2007). The stroop interference effect is the time difference recorded in reaction times naming the colour of the ink in congruent and incongruent trials.
A study by Kuhl & Kazen (1999) explored the effect of emotions and intentions on the stroop interference effect. They hypothesized cognitive interference could be reduced by both participants intentions and inducing PA. A stroop task was used containing three types of priming words (positive, negative and neutral). Researchers revealed the stroop interference could be expunged by the introduction of PA provided two tasks per trial were performed.
Kuhl & Kazen concluded that memorizing intentions for a second task alongside PA induction removed interference in the initial task.
Links could be drawn between caffeine and PA. Like heroin and cocaine caffeine increases dopamine (DA) by inhibiting reuptake, researchers believe this may explain the addictive nature of caffeine. Caffeine blocks adenosine receptors which increases alertness and releases adrenaline into the body manipulating the production of DA making a person feel good.
A neuropsychological theory devised by Ashby, Valentin & Turken (2002) suggests cognitive behavioural effects of PA is mediated by DA release into the anterior cingulate cortex. This association is supported by Beaty (1995) who revealed increased dopaminergic activity after ingesting caffeine increased mood.
Based on the research presented there will be three experimental hypotheses. First, participants in the high caffeine condition will perform better on the stroop task than the low caffeine condition, Second, participants with high PA scores will perform better on the stroop task than those with low PA scores, finally participants with high caffeine levels and high PA scores will show an interaction effect on stroop bias scores.
Method:
Design: This study used a 2×2 mixed factorial double blind design with two independent variables (IV’s), the caffeine dosage and PA. The caffeine dose was within participants design and PA was between participants design. Each IV had two levels; the caffeine dose was either 200mg or 400mg and PA was low or high. PA was not manipulated during the study however caffeine may have inadvertently affected this. The dependent variable (DV), stroop bias scores, was measured by subtracting the reaction times (in milliseconds) of the congruent trials from the incongruent trials.
Participants: 243 second year Coventry University Students created an opportunity sample during science of the mind workshops. Participants were allocated to the low or high PA conditions based on their score on the shortened version of the Positive and Negative Affect Schedule (PANAS). The low PA condition, categorised by a score between 12 and 21, contained 159 participants (M=17.31, SD= 2.26). The condition consisted of 33 males and 126 females (Mage=20.31 years, SD= 2.53). The high PA condition, categorised by a score between 22 and 36, contained 84 participants (M=26.50, SD= 4.42). The condition consisted of 14 males and 70 females (Mage= 19.83 years, SD= 1.17).
Materials: Participants were given an information sheet containing ethical consideration and the requirements of the study. Participants required a signed consent form with student ID number to participate. Each participant was given a practice Stroop task containing four trials with twelve words. Each trial contained the words red, blue, green and yellow. The first trial contained the words written in black ink. Trials two and four contained incongruent words to colours, trial three contained congruent words to colours.
Created by John Ridley Stroop in 1935, the Stroop tests aim to test cognitive performance, Stroop’s first experiment involved 10 words in 10 rows and 10 columns. All words were written in black, the correct response for the word blue was to name the word not the colour it was written in. In Stroop’s second experiment participants were required to name the colours of incongruent colour word combinations, for example, the correct response for the word blue written in yellow, is yellow. Stroop found colour naming to be dramatically slower for incompatible words.
Participants received a shortened version on the PANAS containing 20 items describing positive and negative feelings and emotions, for example excited or distressed. This was approximately one side of A4 in length and presented in Times New Roman font size 12.
Participants were also required to take part in a computerised Stroop task containing 20 practice trials and 120 experimental trials. The stroop task word stimuli were red, blue, green and yellow, presented in Arial font size 20 for a duration of 20 milliseconds (MS). Responses were recorded by timing the speed participants pressed either R, G, B or Y on a standard QWERTY keyboard.
Finally, participants were presented with a debriefing quiz, approximately one A4 size in length containing questions such as, what was the IV? And what were the main aims of the study?.
Procedure: Data was collected during two, two hour second year workshops. The ethics were approved by the Coventry University ethics committee prior to testing. In workshop one, participants were shown exclusion criteria and instructed to abstain from taking part if they suffered from heart problems, seizures, anxiety, diabetes or caffeine sensitivity (do not usually consume caffeine). If inclusion criteria was met, participants read the information sheet and provided written informed consent. Participants first took part as Group A, taking one 200mg caffeine tablet and waiting 20 minutes, timed by the seminar leader, during this time participants were allowed to take part in a practice stroop test with a partner prior to beginning the task. Each participant took turns to attempt four trials in order to observe the stroop effect.
After 20 minutes, participants were required to fill in an online version of the PANAS. Responses were recorded using a 5 point Likert scale (1= very slightly/ not at all, 5= extremely). On completion of the PANAS, participants began the Stroop task. Firstly, participants completed 20 practice trials before undertaking the 120 experimental trials, pressing R, B, G and Y in rapid response to the stimuli. This was completed individually and in silence.
In workshop two, participants took part as Group B, ingesting two 200mg caffeine tablets totalling 400mg of caffeine, then instructed to wait 20 minutes before continuing with the experiment. During this time participants were given a paper version of the shortened PANAS to complete. After 20 minutes the participants once again took part in the online PANAS and completed 20 practice tests and 120 experimental trials individually and in silence. Finally, participants were debriefed and informed of the caffeine level manipulation and the true aims of the study.
Results:
Table 1: Means (and standard deviations) of Stroop Reaction times (s) in the low and high PA groups with low or high (200mg, 400mg) Caffeine consumption.
PA group 200mg Caffeine 400mg Caffeine
Mean SD Mean SD
Low PA 3.35 5.38 2.10 5.43
High PA 2.79 4.12 2.50 5.73
A 2×2 mixed factorial analysis of variance (ANOVA) was conducted with an alpha value of 0.05. As seen in table 1, participants who scored high on the PA scale and ingested 400 mg of caffeine had only slightly quicker reaction times than participants in the same group who ingested only 200mg, however this was not quicker than participants in the low PA group who ingested 400 mg of caffeine.
After meeting all assumptions and sphericity was assumed, the ANOVA reported no significant effect of caffeine on stroop bias scores, F (1,241) = 1.92, p= .67. Additionally, there was no significant difference in bias scores between the high and low PA groups (p>.05), therefore the ANOVA showed no significant effect of PA score on stroop bias scores, F (1,241) = .047, p= .829. Finally, no significant interaction effect of caffeine and PA on stroop bias scores was recorded, F (1,241) = 6.84, p= .409.
Discussion:
Contrary to all hypotheses, the results showed no significant effects of caffeine on Stroop bias scores, supporting results found by Edwards et al (1996). No significant effect of PA on Stroop bias scores, contravening results found by Kuhl & Kazen(1999) and finally no significant interaction effects of caffeine and PA on Stroop bias scores suggesting the interaction is flawed, however further research is needed.
The design of this study leaves limitations to be considered. The study used a within and a between participants design. The within-Participants design may have issues including practice effects, similar to the results found by Edwards et al (1996). Participants may have improved through practice or lost interest due to repetition during the experimental trials either could mask the effects of Caffeine on participants. Attempts were made to control practice effects, each participant was given the opportunity to practice the stroop task before completing the experimental trials.
The between-participants design (the PA score) was limited by the PANAS being a self-report questionnaire, which is open to demand characteristics and observer bias. The nature of the questions lend themselves to bias as participants answers may reflect their perception of what the researcher seeks. A double blind control was attempted, neither researcher nor participants knew the dosage of caffeine administered.
Peak caffeine absorption levels occur between 30 and 60 minutes after ingestion (Jacobson & Kulling, 1989) Participants were tested 20 minutes after ingestion so absorption was incomplete. The absence of a control group makes measurement of the true effect of caffeine and PA on Stroop bias scores impossible, there was also no control for caffeine consumed before testing.
Increasing absorption time to 60 minutes, controlling for caffeine tolerance, caffeine consumption 24 hours prior to commencement of the stroop task and using a control group would improve reliability and validity of results significantly.
Despite having found no significant effects on stroop task performance caffeine and PA still have practical applications. Caffeine, when paired with aspirin provides effective pain relief for ailments such as migraines and stomach aches. PA can be measured by patients lift in mood and ability to continue their daily routine as pain is relieved.
To conclude, although no significant effects were found in this study, it remains unclear to what extent both caffeine and PA impact on cognition. Due to the prolific use of caffeine in today’s diet society would benefit from improved research into Caffeine and PA on cognition.
References
J. Kuhl, J., Kazén, M., (1999). Volitional facilitation of difficult intentions: Joint activation of intention memory and positive affect removes stroop interference. Journal of Experimental Psychology: General, 128, pp. 382–399.
G. Rowe, J.B. Hirsh, R.A. Andersen (2007). Positive affect increases the breadth of attentional selection. Proceedings of the National Academy of Sciences, 104 (1), pp. 383–388
Van der Stigchel,S., Imants, P., Ridderinkhof, R.K., (2011). Positive affect increases cognitive control in the antisaccade task. Brain and Cognition, 75 (2), pp.177-181.
Snyder & Pamela, 1983)
Jacobson & Kulling, 1989
Edwards et al, (1996