Can more coffee make us smarter?
The effect of coffee consumption and priming with
coffee on math performance
Master thesis, Msc. Marketing Management
University of Groningen - Faculty of Economics and Business
Abstract
The present research tested whether coffee consumption and priming influence problem solving abilities, more specific math performance. Three hypotheses were drawn following the appropriate literature, two for the main effects and one for the interaction effect. Participants were approached during different times of the day to account for dosage and then primed with coffee consumption or a neutral prime before making a math test. The math test was analyzed on correct answers and time it took to complete the test. Results show that the two drawn hypotheses regarding the main effects were rejected as no significant results were found. A marginally significant cross-over interaction effect was found on the time it took participants to complete the math test. The current findings suggest that math performance might not improve following the consumption of coffee or priming, but that a high dosage of coffee consumption together with a neutral prime, or a low coffee dosage combined with the coffee prime can lead to faster math performance. A possible explanation can be that a high dosage of coffee and being primed with coffee consumption can lead a person to over think math problems and thus take longer to solve them.
Keywords: Coffee, priming, problem solving, math performance, interaction effect
Introduction
Caffeine containing drinks are amongst the most widely consumed drinks in the world, with coffee coming as the main source of caffeine (Brice & Smith, 2002). The habit of drinking coffee can even be traced back as far as 1000 years ago in Ethiopia. However it wasn’t until the year 1819 when caffeine was purely isolated from coffee beans for the first time by the German chemist Friedlieb Ferdinand Runge (Waldvogel, 2003). Ever since that time caffeine consumption has been growing worldwide, the immense popularity can possibly be taken from the idea that caffeine is an easy and harmless psychostimulant which can increase wakefulness, concentration or enhance physical performance (Rogers et al. 2013, Mitchell et al, 2014). For this reason caffeine is the most used psychoactive stimulant worldwide (James, 1997). And in terms of caffeine intake, the Netherlands ranks number 1, with an intake of 414 mg of caffeine, per person, per day (Fredholm et al. 1999)
Therefore it is not surprising to see many research papers focusing on caffeine, coffee and its effects on both social and non-social behavior, while others just want to find a reliable way to determine how much caffeine different beverages contain (Sereshti & Samadi, 2014). The general estimate is that a normal cup of coffee contains between 80 to 150 mg of caffeine (Gupta & Gajilan, 2007) In the current research, the focus will be on the effect that different dosages of coffee have on the cognitive performance, or the ability to solve problems specifically.
The question whether caffeine can influence your mental state and cognitive performance has been around for a long time already. Holck (1933) was one of the first to test the effect of caffeine on problem solving skills. And since then, a large amount of articles followed that discuss the effect that caffeine can have on a person’s cognitive abilities. A more detailed discussion of this will follow in the literature review part later. A novel fact of this paper is the fact that, unlike most articles, only coffee will be used as a source of caffeine.
coffee he or she will actually feel like they have consumed coffee. Whereas the effect of caffeine on cognitive performance has been the subject of many research papers and it has been tested extensively already, to current knowledge, coffee consumption has not been used as a prime yet. This can be of importance because, as stated, priming can make people behave accordingly, which means for example that people could feel more active and awake without even actually having consumed coffee. Whether this is actually true will be tested in the current paper. Thus priming with coffee consumption will be used in the present paper as an independent variable, together with the dosage of coffee that has been consumed. Both these variables will be tested on problem solving capabilities.
Furthermore, the interaction effect between priming with coffee consumption and coffee dosage will be tested, meaning that the effect that coffee consumption has on problem solving will be greater when someone is primed with coffee, compared to not being primed with coffee consumption. This is of relevance since the research of Hicks et al. (2009) showed that priming with a substance (marijuana in their paper) can influence cognitive performance, however it only influences people that actually had smoked marijuana. Thus in this paper it will be interesting to discover whether priming has an effect on problem solving on its own, only has an effect through interaction with coffee consumption, or perhaps if it has both a main and an interaction effect. How this will be tested will be further explained in the methodology section.
This paper will be structured as followed: First a literature review will focus in more detail on problem solving and on priming and several hypothesis will be drawn. Secondly, the methodology behind the study will be explained. Next, the results from the study will be give. Lastly, these results will be discussed, together with limitations and possibilities for further research.
Literature Review
considered as a single, isolated process. But instead as a process that interacts with many other cognitive processes like learning, reasoning and decision making. Therefore, for this paper the effects of coffee on problem solving and cognitive performance in general where investigated. To give a clear definition of problem solving Wang & Chiew (2010) are cited:
“Problem solving is a cognitive process of the brain that searches or infers a solution for a given problem in the form of a set of paths to reach a set of expected goals”
As previously said, physical performance can be increased by consuming caffeine. However, James (2014) states that, if used on a normal dietary dose, cognitive performance can also be enhanced by caffeine. In 1933 Holck was one of the first to test this. His research was focused on solving chess related problems and he found that participants that were injected with caffeine solved 7% more problems than those that were injected with a saline solution. The participants in his study were unaware of which solution they were injected with. Though this is an old paper, the findings by Holck (1933) are statistically proven by Warburton (1995). He tested the effect of low and medium doses of caffeine (75 & 150mg) on cognition, among which problem solving. To test problem solving capabilities a Semantic Verification Task (Baddeley 1968) was used. The results showed that problem solving significantly improved after caffeine intake.
lead to worse problem solving abilities, higher caffeine dosage in this case means more than 1 cup (80 to 150mg), it needs to be mentioned here however, that this paper deals with energy drinks and not with coffee.
This idea can be further expanded by following the study from Rogers et al. (2013) as in their study they did find that caffeine positively influenced motor performance, they found evidence that the mental performance (making less errors in choice reaction time or in memory tasks) did not improve amongst medium-high caffeine (250mg) users. To cite from their paper:
“Caffeine fails to make medium-high caffeine consumers ‘smarter’ because, due to tolerance to the effects of caffeine on sleepiness/wakefulness, they gain no net increase in mental
alertness from their habit”
The outcomes of these studies can be contested however since they deal with caffeine-naïve participants (low, or non-consumers of caffeine), and participants that are already experiencing withdrawal effects because participants were asked to abstain from their usual caffeine intake overnight (James, 2014). Interestingly enough Hewlett & Smith (2006) followed a similar approach from the one done by Rogers et al. (2013) though their study shows that caffeine does indeed results in a significant improvement of performance. Borota et al. (2014) also tested the effect that caffeine has on learning and memorization. They performed 2 studies, one with a placebo and 200mg of caffeine, and the second study had the placebo and 3 different doses of caffeine (100, 200 & 300mg). Their results show that a dosage of 200mg caffeine had a significantly higher level of performance than both the placebo and the group that received 100mg of caffeine. The 300mg group also performed better than the placebo and 100mg group, but not significantly different from the group that received 200mg of caffeine. Thus, this could lead to believe that a high dosage of coffee, compared to a low dosage of coffee could lead to higher problem solving performance, considering that a low dosage of coffee means 1 cup of coffee (between 80 - 150mg of caffeine (Gupta & Gajilan, 2007)).
accuracy. The results from their study show that participants that received about 3mg/kg of caffeine (approximately 2 cups of coffee) outperformed the group that did not receive any caffeine. However, they outperformed the group receiving 5mg/kg of caffeine as well, which suggest that while a higher dose of caffeine will increase learning & problem solving capabilities compared to a low consumption of caffeine, still there is a limit to the dosage of caffeine. Bättig & Buzzi (1986) have tested the influence of different doses of caffeine on cognitive performance as well, their doses can both be considered as high doses, being 150mg and 450mg of caffeine. What they found was that in both cases a significant increase in performance could be observed compared with no caffeine, or decaffeinated coffee intake, meaning that in order to improve problem solving, a high dosage of caffeine would be best.
Shukitt-Hale et al. (2013) found however that it is coffee and not only caffeine that has a positive effect on cognition. They also tested different doses of caffeine, and it was discovered that the higher dose resulted in the best results on cognitive performance. Their subsequent study showed that the better results could not only be attributed to caffeine, but to other components of coffee as well. Despite the fact that there are some mixed results regarding the effects of coffee and caffeine on problem solving a hypothesis will be posited. The reason for the positive relation in the stated hypothesis is due to the fact that the majority of papers cited above find that a higher coffee dosage generally leads to better problem solving skills. The mentioned papers that contest this view either do not deal with coffee, but with energy drinks (Trunzo et al., 2014), or they do not mention a negative effect, but just no effect at all for a higher dosage of coffee (Rogers et al., 2013) Therefore, taking the previously mentioned into account the following hypothesis will be posited, whereby problem solving abilities will be tested by using a math test in this paper, the exact procedure of this is explained in the methodology section:
H1: A high dosage of coffee (2 cups of coffee or more) will result in higher problem solving capabilities.
Nonconscious priming can either be done subliminally, which means the person being primed is not aware of the prime. Or supraliminally, which means that the person is aware of the prime, but not of the influence that it might have (Bargh, 2002). Both of these primes have effect on motivations, behavior and judgments (Bargh, 2002).
Ciani & Sheldon (2010) discovered in their study that problem solving performance can be affected by priming as well. In their study a group of university students were asked to do an analogy test and an anagram test, however, prior to the test the participants were randomly assigned to a letter, being either A (having a positive meaning) and F (connected to failure). The results showed that students that were exposed to the letter A significantly improved performance compared to being showed the letter F. Lang & Lang (2010) found in their research as well that it is possible to improve problem solving by priming. Their study focused on test-anxious students, who significantly performed better on tests if they had been primed with competence by thinking of a person they deemed as being very successful. In the current paper, however, people will be primed with consumption, rather than with a specific word or trait. Priming with consumption has been tested previously as well, Van Koningsbruggen and Stroebe (2011) for example found in their research that priming with alcohol can influence cognitive tasks, even if participants did not consume alcohol at all. Participants in their research performed less on a math test if they were previously primed with alcohol related words, since participants believed that alcohol is negatively related to cognitive performance. Hicks et al. (2009) found a similar result in their paper. They tested the priming effect of marijuana and found that if participants expected that marijuana would lead to cognitive and behavioral impairment, they would perform worse on such tasks after being primed with marijuana cues.
Thus, since it is expected that coffee positively influences problem solving capabilities it can be expected that being primed with coffee consumption will also result in better problem solving capabilities, based on aforementioned research. Therefore the following hypothesis will be posited:
Thirdly, it will be investigated whether an interaction effect exist. In other words, it will be tested to see if the effect that coffee has on problem solving will be moderated by being primed with coffee consumption.
Anderson & Horne (2008) Found that it is possible to use a placebo and suggestibility effect to make people aware of the effects of coffee, in their case it was specifically aimed at testing the sleepiness and reaction time of participants in their study. In their study the placebo group was informed that they would receive a highly energizing form of coffee and they were told that it should start working within 15 minutes. Finally, before the actual study they were informed that that are supposed to feel more energized now, in contrast to the control group, who were only asked how they were doing. This study showed that receiving the verbal cues significantly improved reaction time amongst participants.
An actual interaction effect was tested and found by Hicks et al. (2009). They performed 2 sample studies where participants were primed with marijuana. The first sample study showed that priming influences the performance on a timed math test, however it only affected those participants that thought beforehand that marijuana would actually lead to worse results on such a test. In other words, participants that believed that marijuana would lead to poorer results actually performed worse on a math test when primed with marijuana, compared to participants who believed marijuana would not affect them. In their second sample study this was tested further when they found that participants who did not smoke marijuana in the past year were not influenced by the prime, but those participants that had recent experience with marijuana did perform significantly worse on the math test after being primed with marijuana.
Taking these studies into account it can be expected that priming with coffee consumption can make people experience the effects of coffee, without actually having it. And that an interaction effect can be expected as well, meaning that the effect that coffee dosage would have on problem solving will be even bigger when participants are primed with coffee consumption compared to when they are not primed with coffee consumption. Put differently, since it is expected that a high dosage of coffee will lead to better problem solving, the interaction with the prime should lead to an increased effect of the dosage on problem solving (meaning, even better problem solving). In comparison, participants that were not primed with coffee consumption, but only consumed a high dosage of coffee will only show the expected effect of improved problem solving. Thus, the following hypothesis will be tested:
H3: The effect of coffee dosage on problem solving abilities will be greater for participants that have been primed with coffee consumption.
These hypotheses lead to the following conceptual model:
Methodology
The following section will provide a detailed overview of the methods used in order to test the hypothesis. A positive effect was predicted for high coffee consumption of problem solving performance, with this effect becoming even stronger when participants were also primed with coffee consumption.
Participants
In total 100 randomly selected Dutch participants (mainly students) (44 female, 56 male; Mage = 24.36 years, SDage = 8.02) participated voluntarily. The study used a 2 (low
dosage vs. high dosage) X 2 (primed with coffee consumption vs. control prime) between-subjects factorial design. All participants were randomly approached during a four-week time span, between Monday and Friday and it was done in and around several buildings at the Zernike campus of the Rijksuniversiteit Groningen . Participants were either primed with coffee consumption, or primed with the control cue. All participants were approached at the same place to prevent biased results. This location was chosen for the large amount of both students and professors that were either having a coffee or lunch, or who were simply meeting one another. Due to the 2 X 2 factorial design, 4 cells appear in the study, thus to reach an adequate amount of 25 unique respondents per cell a total of 100 participants were required in the total study.
Procedure
An examiner approached potential participants and asked them if they wanted to partake in an experiment that was introduced as being part of a master thesis for the Faculty of Economics and Business, however the exact purpose of the research was not explained. Participants were informed that they would be asked several questions and that they had to perform a math related task. Participants were approached either early in the morning (between 8:30 and 11:30 a.m.) to meet the low dosage requirement, or they were approached after midday (between 12:00 and 5:30 p.m.) to meet the high dosage requirement. If they agreed to participate to the research they were first given a short introduction that stated that they would be given several unrelated questions and tasks on language and reasoning, this was done in order for them to not guess the true meaning behind the experiment. They were also ensured that their participation would be fully anonymous. For the full written introduction please see Appendix 1.
was met the participants were given a timed math test to fill in (Appendix 5). Following the math test participants were asked several question regarding their coffee consumption (e.g. amount of coffee they had today, when they had their last coffee, their current need state and the average consumption for each day of the week (Appendix 7)) and they were asked to fill out an 18-item need for cognition scale (Cacioppo et al., 1984)(Appendix 6) that was used as a control variable personality scale for problem solving abilities. All sessions were coded. After the interview was completed participants were debriefed about the purpose of the experiment and if they would like to be followed-up with results (Appendix 2).
Independent variables
Coffee dosage
The effect of coffee and different doses of coffee on problem solving abilities has been tested and both positive (Shukitt-Hale et al., 2013, Hewlett & Smith, 2006, Borota et al., 2014, Young & Molesworth, 2011) and negative (Trunzo et al., 2014, Rogers et al., 2013) effects have been found. However to determine what can be seen as a low dose of coffee and what can be seen as high it is important to consider the amount of caffeine in a regular cup of coffee. In general, the papers studied used between 0mg of caffeine up to 100mg of caffeine as a low dose. Following Gupta & Gajilan (2007) one cup of coffee generally contains between 80 – 150mg caffeine. It was assumed that in the morning participants did not drink more than one cup of coffee yet, therefore, participants that were approached in the early morning (between 8.:0 and 11:30 a.m.) were assigned to the low dosage condition. Furthermore, it was assumed that after midday (between 12:00 and 5:30 p.m.), participants would have had their second cup of coffee or more already and thus participants that were approached after during this time were assigned to the high dosage condition. This distinction was made to come to a completely random experiment, actual coffee consumption was asked afterwards and was used as a control condition.
Priming with consumption
control cue, in this case the last time they took the bus. To vary between these two conditions participants had to write down answers to three questions about the last time they drank multiple cups of coffee in a short time, or similarly, they had to write down answers to three questions related to the last time they took the bus. As stated previously, the first question was to briefly describe the last time the participant drank multiple cups of coffee/took the bus, the second question asked the participants to describe the situation when this happened and the third question asked them to describe their feelings at that moment (Appendix 3 & 4). By just asking participants these simple questions and having them write it down non-conscious behavior was activated, which makes the participants behave accordingly (Bargh, 2002).
Dependent variable
The main purpose of the experiment was to find out if there is an interaction effect of consumption dosage and priming with consumption on problem solving. As stated, problem solving is defined in this paper using the definition from Wang & Chiew (2010):
“Problem solving is a cognitive process of the brain that searches or infers a solution for a given problem in the form of a set of paths to reach a set of expected goals”
In order to test this a math test was used, this approach comes from Van Koningsbruggen and Stroebe (2011), who found in their research that alcohol consumption can influence problem solving, which they tested using 30 moderately complex equations. This method was adopted from the paper from Van Dillen & Kool (2007). The test used by Van Koningsbruggen and Stroebe measures the baseline math performance. Therefore, as a suitable variant of this math test, in this paper the “Rekentest referentieniveau 2F1” was used (Appendix 5). This test was developed by the Dutch Ministry of Education, Culture and Science and it is meant to test the average math level of a person. Test 2F was specifically chosen because, just like the test from van Koningsbruggen and Stroebe, it tests baseline math performance. The test consists of 13 questions, asked in Dutch. Both the amount of correct questions and the speed of answering were used to measure the possible effects.
1
Control variables
After the math test participants also filled in the need for cognition scale, this scale was taken from Cacioppo et al. (1984) who describe need for cognition as: “An individual’s tendency to engage in and enjoy effortful cognitive endeavors”. In other words, it can be expected that participants that score higher on this scale perform better on math performance as well. The need for cognition scale consists of 18 items that were answered on a 5-point likert scale, and it was found to be highly reliable with α = 0.87 (Appendix 6)., reverse scoring was used on 9 items and this was accounted for. Afterwards the scores from all 18 items were added together to get the total need for cognition score that was used in the analysis (Mneed_for_cognition_total = 63.36, SDneed_for_cognition_total = 9.31).
Furthermore, participants were asked to fill in a questionnaire consisting of the participants age, gender, average weekly coffee consumption, the current need state of the participant and the impact expectancy which were used as control variables. Need state was measured using one item: “On a scale from 1 to 7, how much do you want a cup of coffee right now?”, with 7 meaning that the participant was in a high need for coffee. Impact expectancy was as well measured using one item: “If I drink two cups of coffee in one hour, my ability in solving choice and skill tasks will be?”, this was measured on a 9-point likert scale, from 1 (much worse) to 9 (much better). For coffee consumption participants were asked for their average coffee consumption for each day of the week and the total of this was divided by 7 to come to an average coffee consumption. (Appendix 7 contains the full questionnaire)
Results
and priming, the analysis showed that a high dose of coffee combined with being primed with coffee consumption did not lead to significantly better math performance (F(1,96) = 0.01, p = 0.90). These results could already suggest that all 3 hypotheses can be rejected.
However, the time taken to finish the math test was also measured, this again showed no significant main effect for both dosage (F(1,96) = 0.47, p = 0.49) and priming (F(1,96) = 0.59, p = 0.59). Though here there is a marginally significant interaction effect with (F(1,96) = 3.89, p = 0.05). As figure 1 shows, there is a cross-over interaction effect, which in this case means that participants that had a high dose of coffee take a shorter amount of
time to finish the math test when they are not primed with coffee, but when they had a high dose of coffee and when they were primed with coffee they actually took a longer time to finish (Mhigh dosage X primed = 15:56 minutes, SDhigh dosage X primed = 4:44 minutes versus Mhigh dosage X not primed = 14:43 minutes, SDhigh dosage X not primed = 4:16 minutes). On the other hand however, if a participant only had a low dose of coffee they finish the math test faster when they are primed with coffee consumption. However, participants took a longer time again to finish the test when they had a low dose of coffee and were not primed with coffee consumption (Mlow dosage X primed = 14:41 minutes, SDlow dosage X primed = 4:25 minutes versus Mlow dosage X not primed = 16:59 minutes, SDlow dosage X not primed = 3:52 minutes)
Before completely rejecting the first 2 hypotheses, regarding the main effects of dosage and priming on math performance, another analysis was performed. This time an ANCOVA was used to control for several factors. First an ANCOVA was run with the amount of correctly solved answers as dependent variable, with each control variable (gender, age, average coffee consumption, the need state of the participant, the impact expectancy and the need for cognition) entered at once. However, again there were no significant effects for both main effects of dosage (F(1,90) = 0.03, p = 0.85) and priming (F(1,90) = 0.12, p = 0.73 and the interaction effect (F(1,90) = 0.00, p = 0.97) on correctly answered math problems. None of the covariates were significant as well. Thus from this it can be concluded that neither drinking 2 cups of coffee or more, nor priming with coffee consumption results in better math performance in terms of correctly answered questions.
Lastly, an ANCOVA was performed with the amount of time taken to finish the math test as dependent variable and the aforementioned control variables as covariates. Again, the covariates were all entered at the same time into the model. The results show that there is no significant main effect for both dosage (F(1,90) = 0.13, p = 0.72) and priming (F(1,90) = 0.32, p = 0.57). Furthermore, there was also no longer a marginally significant cross-over interaction effect (F(1,90) = 2.45, p = 0.12). The possible reason for the lack of a significant interaction effect can be explained by the gender of the participants which is highly significant at (F(1,90) = 7.50, p = 0.01). In other words, the reason why a significant interaction effect was found before can be explained by the fact that men and women differ in the amount of time it takes them to finish the math test (Mmale = 14:41 minutes, SDmale = 3:54 minutes versus Mfemale = 16:49 minutes, SDfemale = 4:30 minutes). Furthermore, the need for cognition scale also showed to be marginally significant (F(1,90) = 3.43, p = 0.07), a possible reason for this is that persons who score high on the need for cognition scale (e.g. who enjoy engaging in effortful cognitive behavior) might enjoy to take their time in solving math problems compared to participants scoring lower on need for cognition. None of the other covariates were significant.
participants to finish the math test, not on correctly answered questions. And this interaction effect only exist when the covariates are not being controlled for.
Discussion
Results from the study show that contrary to predictions, a higher dosage of coffee and being primed with coffee consumption do not significantly affect the performance of participants on a math test. Neither in the amount of correctly answered questions, nor in the time it took them to complete the test.
Rogers et al.’s (2013), since they also found that caffeine consumption did not affect the number of errors made in a memory task, but it did affect choice reaction time, meaning they were faster in making decisions as well. These results can lead to believe that if you are required to make fast decisions it is best to either drink more coffee, or think about more coffee.
What needs to be noted here is that this interaction effect was no longer marginally significant after controlling for age, gender, average coffee consumption, the current need state of the participant, impact expectancy and the need for cognition trait. A possible reason for this and potentially the reason for not finding significant effect on the amount of correct answers can be found in the participant group, since almost all participants were University students it can be argued that these students possess above average problem solving skills compared to for example blue-collar workers. The need for cognition scale should account for this, as it shows how likely people are to engage in, and enjoy effortful cognitive endeavors (Cacioppo et al., 1984). Need for cognition indeed turned out to be marginally significant when it was being controlled for, indicating that participants rating higher on this scale enjoy solving such math problems and they will take their time to be absolutely sure they answered correctly. This view is further strengthened by the fact that when everything was controlled for, except need for cognition, there was a marginally significant interaction effect again (p = 0.9).
For practice, these results mean that, contrary to popular belief, drinking more coffee does not help you to solve problems any better or faster compared to when you would drink less coffee. Furthermore, since priming makes people behave according to the concept of coffee in this case, it means that priming will not result in increased problem solving capabilities. Since this study focused mainly on student participants, it is interesting for them to discover that in order to perform better on exams (another form of problem solving), drinking a lot of coffee is not the optimal solution.
experiment where coffee consumption can be closely administered. In such a lab experiment people that do not consume coffee at all can also be omitted, since the present study featured several participants that never drank coffee, which can also affect the results.
Secondly, the interviews lasted longer than previously expected, thus it is possible that participants lost attention after a certain moment, which can question the reliability of these participants. Further research can possibly feature a similar but somewhat shorter math test. However, most participants indicated that they liked the challenge of the math test and that they did their best to answer everything correctly, therefore there is no reason to believe that participants filled in random answers.
Next, as stated before, the overall group of participants consisted of mainly young students Mage = 24,36, SD = 8,023. Several professors were also interviewed in order to account for this. However, in case of a follow-up lab study, a greater difference in participants can be used to potentially receive more significant results. What needs to be noted is that, despite the lack of incentive to participate in the interviews, all participants were still very willing to participate. What further strengthens the reliability is the fact that the questionnaire was set up in such a way that participants found the interview questions very unrelated, indicating that they could not guess for the results and the presence of possible relationships between variables, and as such behave accordingly to the results.
References
Anderson, C., & Horne, J.A. 2008. Placebo response to caffeine improves reaction time performance in sleepy people. Human Psychopharmacology, 23: 333-336
Baddeley, A.D. 1968. A three-minute reasoning test based on grammatical transformation.
Psychonomic Science, 10(10): 341-342
Bargh, J.A. 2002. Losing consciousness: Automatic influences on consumer judgment, behavior, and motivation. Journal of Consumer Research, 29(2): 280-285.
Bargh, J.A., Chen, M., & Burrows, L. 1996. Automaticity of Social Behavior: Direct Effects of Trait Construct and Stereotype Activation on Action. Journal of Personality and Social Psycholog, 71(2): 230 - 244.
Bättig, K., & Buzzi, R. 1986. Effect of coffee on the speed of subject-paced information processing. Neuropsychobiology, 16(2-3): 126-130
Borota, D., Murray, E., Keceli, G., Chang, A., Watabe, J.M., Ly, M., Toscano, J.P., & Yassa, M.A. 2014. Post-study caffeine administration enhances memory consolidation in humans. Nature
Neuroscience, 17(2): 201-203
Brice, C.F., & Smith, A.P. 2002. Factors associated with caffeine consumption. International
Journal of Food Sciences and Nutrition, 53: 55–64
Ciani, K.D., & Sheldon, K.M. 2010. A versus F: The effects of implicit letter priming on cognitive performance. British Journal of Educational Psychology, 80: 99-119
Fredholm, B.B., Bättig, K., Holmén, J., Nehlig, A., & Zvartan, E.E. 1999. Actions of caffeine in the brain with special reference to factors that contribute to its widespread use. Pharmacological
Reviews, 51: 83-133
Gupta, S. & Gajilan, A. 2007. The caffeine habit. Time 170(14): 62.
Hicks, J.A., Pedersen, S.L., McCarthy, D.M., & Friedman, R.S. (2009). Marijuana primes, marijuana expectancies, and arithmetic efficiency. Journal of Studies on Alcohol and Drugs, 70(3): 391– 399.
Holck, H.G.O. 1933. Effect of caffeine on chess problem solving. Journal of Comparative
Psychology, 15(2): 301-311
James, J.E. 2014. Caffeine and cognitive performance: Persistent methodological challenges in caffeine research. Pharmacology, Biochemistry and Behavior, 124: 117-122
James, J.E. 1997. Understanding caffeine: A biobehavioral analysis. Thousand Oaks, CA: Sage. Lang, J.W.B., & Lang, J. 2010. Priming competence diminishes the link between cognitive test anxiety and test performance: Implications for the interpretation of test scores. Psychological
Science, 21(6): 811-819
Mitchell, D. C., Knight, C. A., Hockenberry, J., Teplansky, R., & Hartman, T. J. 2014. Beverage caffeine intakes in the U.S. Food and Chemical Toxicology 63: 136–142.
Rogers, P.J., Heatherley, S.V., Mullings, E.L., & Smith, J.E. 2013. Faster but not smarter: Effects of caffeine and caffeine withdrawal on alertness and performance. Psychopharmacology, 226: 229-240
Sereshti, H., & Samadi, S. 2014. A rapid and simple determination of caffeine in teas, coffees and eight beverages. Food chemistry, 158: 8-13
Shukitt-Hale, B., Miller, M.G., Chu, Y., Lyle, B.J., & Joseph, J.A. 2013. Coffee, but not caffeine, has positive effects on cognition and psychomotor behavior in aging. American Aging Association, 35: 2183–2192
Trunzo, J.J., Samter, W., Morse, C., McClure, K., Kohn, C., Volkman, J.E., & O’Brien, K. 2014. College students’ use of energy drinks, social problem-solving, and academic performance.
Journal of Psychoactive Drugs, 46(5): 396-401
Waldvogel, S. R. 2003. Caffeine a drug with a surprise. Angewandte Chemie International
Edition, 42: 604–605.
Wang, Y., & Chiew, V. 2010. On the cognitive process of human problem solving. Cognitive
Wang, Y., Wang, Y., Patel, S., & Patel, D. 2006. A layered reference model of the brain (LRMB).
IEEE Transactions on Systems, Man, and Cybernetics, 36(2): 124–133
Warburton, D.M. 1995. Effects of caffeine on cognition and mood without caffeine abstinence.
Psychopharmacology, 119: 66-70
Young, R., & Molesworth, B.R.C. 2011. The Effects of Caffeine on Learning: A Pilot’s perspective.
Appendix 1 – Introduction
Introduction
Het onderzoek bestaat uit enkele ongerelateerde deelstudies naar taal en redeneren en maakt deel uit van het onderzoeksprogramma van verschillende onderzoekers aan de Faculteit Economie en Bedrijfskunde aan de Rijksuniversiteit Groningen. Deelname is vanzelfsprekend geheel anoniem en je gegevens zullen uitsluitend vertrouwelijk worden gebruikt voor dit onderzoek en niet verder worden verspreid. Het onderzoek zal slechts enkele minuten van je tijd vergen. Werk a.u.b. zelfstandig aan het onderzoek en overleg niet met anderen.
Appendix 2 - Debriefing
Appendix 3 – Priming with coffee consumption
Beschrijf hieronder in enkele zinnen een situatie uit je eigen leven waarin je in korte tijd meerdere koppen koffie hebt gedronken.
Beschrijf hieronder wat er precies gebeurde in die situatie waarin je in korte tijd meerdere koppen koffie hebt gedronken:
Beschrijf hieronder welke gedachten en gevoelens er door je heen gingen toen je in korte tijd meerdere koppen koffie hebt gedronken:
Appendix 4 – Control prime, bus
Beschrijf hieronder in enkele zinnen de laatste keer dat je de bus hebt genomen.
Beschrijf hieronder wat er precies gebeurde in die situatie waarin je de bus hebt genomen:
Beschrijf hieronder welke gedachten en gevoelens er door je heen gingen terwijl je de bus nam:
Appendix 5 – Rekentest referentieniveau (Correct answers are in bold and underlined)
Nu gaat u enkele opgaves maken die het redeneren in kaart brengt.
1. U heeft een privérekening bij een bank en u kunt € 1000 "rood staan". U heeft € 198 tekort op deze rekening. Welk bedrag kunt u nu ten hoogste pinnen van deze privérekening?
a. € 0
b. € 802
c. € 1000 d. € 1198
2. Een woonkamer in de vorm van een rechthoek is 6 meter breed en 7 meter lang en moet bekleed worden met tapijt van 4 meter breed. Hoeveel meter van dit tapijt moet u minimaal aanschaffen om de woonkamer volledig te bekleden en u snijresten niet wilt gebruiken?
a. 6 meter b. 7 meter
c. 12 meter
d. 14 meter
3. U reist op woensdag met de trein van Hengelo naar Deventer en weer terug. De heenreis vindt plaats voor negen uur 's ochtends en de terugreis in de middag. U heeft een NS Voordeelurenkaart waarmee u elke werkdag na negen uur met 40% korting kunt reizen. Een retourbiljet kost 1,75 keer de prijs van een enkele reis. Wat is in dit geval voor u het goedkoopst?
a. U koopt een retourbiljet Hengelo – Deventer tegen het normale tarief.
b. U koopt een enkele reis Hengelo – Deventer tegen het normale tarief en een enkele reis Deventer – Hengelo tegen het voordeeltarief.
c. Beide mogelijkheden zijn even duur.
4. De zomer is bijna voorbij en enkele kledingzaken in uw woonplaats ruimen hun zomervoorraad op. Bij kledingwinkel Fashion geldt: "3 Halen, 1 Betalen!" en bij Modeshopper: "U betaalt de helft van de helft!". U wilt drie dezelfde zomerartikelen kopen, die elk even duur zijn. Bij welke winkel krijgt u relatief de hoogste korting?
a. Fashion
b. Modeshopper
c. Beide winkels geven dezelfde korting
5. Uw auto loopt 1 op 14 (dat wil zeggen: u heeft 1 liter benzine nodig om 14 kilometer te rijden). De benzinetank heeft ruimte voor 60 liter. U heeft een afspraak op 245 kilometer van uw huis. Bij vertrek van uw huis is uw tank nog maar een kwart vol. Hoeveel liter moet u minimaal tanken om thuis te komen?
a. 10 liter b. 14 liter
6. Omdat u bezorgd bent dat uw kinderen te veel suiker binnen krijgen, bestudeert u in de supermarkt nauwkeurig de suikergehalten van een aantal merken cola. De resultaten zijn:
Merk Hoeveelheid suiker
Popsa 200 gram per fles van 1,5 liter Cila 140 gram per fles van 1 liter Flizz 80 gram per fles van 0,5 liter Coolcoke 50 gram per fles van 0,33 liter Welk merk cola bevat het laagste suikergehalte?
a. Popsa
b. Cila c. Flizz d. Coolcoke
7. Er is een supermarktoorlog gaande in uw woonplaats. Een supermarkt geeft 20% korting op een kilo gehakt. Van sommige verpakkingen gehakt is de uiterste houdbaarheidsdatum bijna verlopen en daarom geeft de supermarkt op deze afgeprijsde kilo's nog eens 50% extra korting. Hoeveel korting krijgt u in totaal op een kilo gehakt?
a. 40% b. 50%
c. 60%
d. 70%
8. U gaat in uw tuin een vierkante vijver aanleggen van 60 cm diep en zijden 160 bij 160 cm. U voert al het zand uit de vijver af met een kruiwagen naar een hoek van uw tuin. Hoe vaak moet u ongeveer met een volle kruiwagen naar de hoek van uw tuin lopen als u weet dat een kuub zand ongeveer 13 kruiwagens omvat?
a. 2 keer b. 5 keer
c. 20 keer
d. 50 keer
9. U wilt een terras aanleggen in uw tuin in de vorm van een rechthoek met behulp van 300 rechthoekige siertegels, die elk dezelfde afmeting hebben. U legt uw voorstel voor aan uw partner, die u de suggestie doet het terras twee keer zo lang en twee keer zo breed te maken. Hoeveel siertegels heeft u in dat geval nodig?
a. 600 b. 900
10. Digitale gegevens zoals documenten, foto's, muziekfragmenten en films kunnen worden opgeslagen op CD-roms en DVD's. De ruimte die deze schijven bieden wordt uitgedrukt in een aantal bytes. Omdat deze opslagruimte een groot aantal bytes omvat, wordt vaak gesproken van megabytes of gigabytes. Een CD-rom heeft een opslagruimte van 700
megabyte. Een DVD heeft een opslagruimte van 4,7 gigabyte. Hoeveel CD's passen er op een DVD?
a. ongeveer 0,67
b. ongeveer 6,7
c. ongeveer 67 d. ongeveer 670
11. De economische recessie komt onder andere tot uitdrukking in het aantal openstaande vacatures bij bedrijven en instellingen. Het Centraal Bureau voor de Statistiek heeft onderstaande grafiek gemaakt voor het aantal openstaande vacatures in een aantal kwartalen vanaf de zomer van 2005.
Wanneer steeg het aantal openstaande vacatures het sterkst? a. van het derde naar het vierde kwartaal in 2005
b. van het tweede naar het derde kwartaal in 2006
c. van het eerste naar het tweede kwartaal in 2007 d. van het derde naar het vierde kwartaal in 2008
12. In 2008 berekende de Belastingdienst de arbeidskorting op de belastingaanslag van iemand jonger dan 57 jaar met een inkomen hoger dan € 8.587 volgens de rekenregel: Neem 12,430% van het deel van het inkomen dat hoger is dan € 8.587 en tel bij de uitkomst € 151 op. Welke formule geeft deze rekenregel correct weer?
a. arbeidskorting = 12,430% van inkomen + € 151
b. arbeidskorting = 12,430% van (inkomen – € 8.587) + € 151
c. arbeidskorting = 12,430% van inkomen - € 8.587
13. De Internationale Schaatsunie ISU publiceert op haar website een overzicht van hoe de schaatsrecords zich in de loop van de tijd ontwikkeld hebben. De onderstaande figuur geeft aan op welk moment welke records behaald zijn door mannen met klapschaatsen op de 10.000 meter lange baan in een indoorbaan in de periode 1998 – 2007. De verschillende recordtijden zijn als punten in de
onderstaande figuur weergegeven.
Wanneer werd het schaatsrecord het meest verbeterd? a. december 2000
b. februari 2002 c. december 2005
d. februari 2007
Appendix 6 – Need for cognition scale
Hieronder staan enkele stellingen m.b.t. redeneren die u moet beantwoorden met een cijfer van 1 tot 5.
1= Geen goede afspiegeling van mij, 2= enigszins geen afspiegeling van mij, 3= neutraal, 4= enigszins een afspiegeling van mij, 5= goede afspiegeling van mij
1. Ik geniet echt van een taak waarin men met nieuwe oplossing voor 1 2 3 4 5 problemen moet komen.
2. Ik ben graag verantwoordelijk voor het regelen van situaties waarbij 1 2 3 4 5 ik veel moet nadenken.
3. Denken is niet bepaald mijn hobby. 1 2 3 4 5 4. Ik doe liever iets waarbij weinig nagedacht hoeft te worden, dan iets
waarbij mijn denkvermogen zeker op de proef wordt gesteld. 1 2 3 4 5 5. Ik denk liever na over kleine dagelijkse zaken dan over lange
termijn zaken. 1 2 3 4 5
6. Ik doe liever dingen die ik beheers en waarbij ik niet zo hoef na te
denken. 1 2 3 4 5 7. Ik verkies ingewikkelde boven eenvoudige problemen. 1 2 3 4 5 8. Ik probeer vooraf situaties te vermijden waarin ik de kans loop diep
over iets te moeten nadenken. 1 2 3 4 5 9. Het geeft me voldoening iets grondig en lang te overdenken. 1 2 3 4 5 10. Ik denk alleen zoveel als nodig is. 1 2 3 4 5 11. Het idee om de top te bereiken door op mijn verstand te vertrouwen
spreekt mij aan. 1 2 3 4 5
12. Ik loop niet zo warm voor het leren van nieuwe manieren om over
dingen na te denken. 1 2 3 4 5
13. Ik houd ervan om in het dagelijkse leven geconfronteerd te worden
met problemen waarvoor ik zelf met een nieuwe oplossing moet 1 2 3 4 5 komen.
14. Het idee om abstract te denken vind ik aantrekkelijk. 1 2 3 4 5 15. Ik verkies een intellectuele, moeilijke en belangrijke taak boven een
taak die niet onbelangrijk is, maar die niet veel nadenken vergt. 1 2 3 4 5 16. Ik voel me eerder opgelucht dan tevreden na het voltooien van een
taak die veel geestelijke inspanning vergde. 1 2 3 4 5 17. Ik vind het voldoende dát iets werkt, hoe of waarom het werkt, kan
me niet schelen. 1 2 3 4 5 18. Ik zit vaak over zaken te peinzen, ook als ze me niet persoonlijk
Appendix 7 – Final questionnaire, including control variables
Geslacht
M F
Leeftijd …………..
Wil je hieronder aangeven hoeveel koppen koffie (dat mag ook espresso, cappuccino, macchiato of een andere koffievariant zijn) je meestal drinkt op
elk van de onderstaande dagen van de week:
Aantal koppen koffie:
Maandag ………… Dinsdag ………… Woensdag ………… Donderdag ………… Vrijdag …….….. Zaterdag ………… Zondag …………
Hoeveel koppen koffie heeft u vandaag gedronken?
Op welk tijdstip heeft u uw laatste kop koffie gedronken? Geef ook de grootte en soort koffie aan
(espresso, latte).
Koppen koffie ……… Hoe laat heeft u uw laatste
kop koffie gedronken?
……….
Grootte (S/M/L) Soort
Op een schaal van 1 tot 7, hoe graag zou u op dit moment een kop koffie willen?
(1= helemaal niet, 7 = heel erg graag)
1 2 3 4 5 6 7
Als ik 2 koppen koffie drink in 1 uur wordt mijn vaardigheid in keuze- en behendigheidsspelletjes: (1 = Veel slechter; 5= Geen invloed; 9= Veel beter)
1 2 3 4 5 6 7 8 9
