Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Water supplementation after dehydration improves judgment and decision-making performance

Water supplementation after dehydration improves judgment and decision-making performance Previous research has shown that dehydration and water supplementation affect mood and cognitive performance in both adults and children on a variety of tasks that assess memory, attention, executive function, and speeded responses. Given the varied effects of water on cognition, this study explored potential effects of water supplementation, hydration status, and thirst on thinking and decision-making tasks. 29 adult participants undertook a battery of cognitive tests on two separate occasions after having fasted from the previous night. On one occasion, they were offered 500 ml of water to drink prior to testing. Measures of urine osmolality confirmed the group-level effectiveness of the dehydration manipulation. Water sup- plementation was found to improve performance on tasks measuring cognitive reflection in judgement and decision-making. This increase in performance was associated with differences in tasks implicated in inhibition processes. Drinking water after a 12-h dehydration period increased performance in judgement and decision-making tasks, and this was not explained by differences in subjective thirst or attentiveness. Introduction negatively affected (Bar-David, Urkin, & Kozminsky, 2005; Fadda et al., 2012). Several studies have found that dehydration affects not only In addition to dehydration studies, the effects of water mood but cognitive performance in both adults and children consumption on cognitive performance have been exam- (see Benton, 2011 for review). For example, Sharma et al. ined separately. Edmonds, Crombie, Ballieux, Gardner, and (1986), who dehydrated their participants to 1%, 2% and 3% Dawkins (2013) found that water consumption resulted in loss of body weight, found that performance on a memory improved performance on a visual attention task at both test and psychomotor stylus test worsened with increasing 20 min and 40 min after water supplementation when com- levels of dehydration. Similarly, exercise-induced dehydra- pared to baseline scores. In similar studies conducted in chil- tion of 2% loss of body weight was found to negatively affect dren, water supplementation was found to improve perfor- performance on memory tasks, mathematical calculation mance on memory and attention tasks (Edmonds & Jeffes, tasks, and executive function tasks (Gopinathan, Pichan, & 2009; Edmonds & Burford, 2009) and on a letter cancella- Sharma, 1988). More recently, Benton, Jenkins, Watkins, tion task, which assesses visual attention (Booth, Taylor, & and Young (2016) reported improved memory and focussed Edmonds, 2012). attention in individuals dehydrated to 1% body loss. Further- Some researches have been conducted in which the effects more, studies investigating the effects of dehydration in chil- of thirst have also been evaluated. In one study, it was found dren have found that their performance on memory tasks is that participants self-reporting as being very thirsty showed a dose-dependent improvement on a sustained attention task when drinking water (Rogers et al., 2001). More surpris- Olivia C. Patsalos and Volker Thoma are co-first authors. ingly, though, participants who reported a low thirst level showed a dose-related impairment in performance on the * Olivia C. Patsalos same task. The role of thirst on cognitive performance was olivia.patsalos@kcl.ac.uk further evidenced in a study by Edmonds et al. (2013), in Section of Eating Disorders, Department of Psychological which it was reported that thirst facilitated performance on Medicine, King’s College London, London SE5 8AZ, UK tasks requiring controlled processing (set shifting). This University of East London, Stratford Campus, Water Lane, study also reported that thirsty participants performed London E15 4LZ, UK Vol.:(0123456789) 1 3 Psychological Research significantly better on a simple reaction time task after Raichle, 1990). We therefore employed a number of execu- drinking water and that this was moderated by participants’ tive function tasks to test possible links between hydra- subjective feelings of thirst. tion and executive functions such as inhibition, set shift- In light of these findings, it is surprising that to the best ing and updating processes (Miyake et al., 2000). Research of our knowledge, no studies have been investigating the has shown that the performance on the CRT and heuristics impact of dehydration and water supplementation on judge- thinking vignettes may also depend on executive functions, ment and decision-making, crucial tasks in everyday life. such as inhibition capability (De Neys & Glumicic, 2008). This is even more surprising, as there are clear links between Specifically, if overcoming automatic thinking patterns executive functions and judgment and decision-making per- (such as the ‘implied’ but incorrect answers in the heuristic formance (Cokely & Kelley, 2009). For example, Toplak, vignettes and the CRT) is due to inhibition (of automatic— West, and Stanovich (2011) found that the cognitive reflec- ‘intuitive’ responses), then we predicted that performance on tion test (CRT) correlates positively with working memory, an alternative choice reaction time task (ChoiceRT) would inhibition, and set-shifting performance. Hence, in the pre- affect the relationship between water supplementation and sent investigation, we have employed tasks that typically tap cognitive reflection scores. This was measured in the cur - judgement and decision-making processes, aiming to capture rent study with the ChoiceRT, which measures Stroop-like any potential effects of water supplementation and thirst on inhibition performance (we used the Cambridge Neuropsy- thinking and reasoning abilities when participants are in a chological Test Automated Battery (CANTAB) software dehydrated state. package (Sahakian & Owen, 1992) to assess cognitive People are faced daily with judgements and decisions performance, see Methods). In addition to inhibition pro- which potentially put great demands on our cognitive pro- cesses, judgement and decision-making performance may cesses. Consequently, people often rely on heuristics (mental also rely on the ability to symbolically manipulate mental shortcuts) that simplify the task at hand—even in cases in representations (see Buckner & Carroll, 2007) and deal with which the task is a relatively simple one to solve (Frederick, calculations (Toplak et al., 2011). Hence, we also employ a 2005). A sizeable body of research suggests that heuristics set-shifting task, the CANTAB Intra-Extra Dimensional Set can lead to systematic deviations from logic, resulting in Shift (IED), which measures cognitive flexibility and the predictable biases and inconsistencies (Kahneman, 2003). ability to keep mental concepts or representations separate. When faced with certain probabilistic judgment problems, Finally, solving judgement vignettes may also require sus- people tend to use heuristics instead of reflective thinking tained attention (see Booth et al., 2012) and regular updat- that resembles an ‘algorithmic’ approach—the considered ing of working memory content (Miyake et al., 2000), as use of rule-based processes that should lead to normatively measured with the CANTAB Rapid Visual Processing task correct outcomes (though see Gigerenzer & Goldstein, 1996 (RVP), in which participants have to monitor a sequence of for a different view of heuristics). This led to a distinction numbers and respond when observing a target sequence of between automatic (heuristic) and reflective (analytic) think - three consecutive numbers. ing, described as ‘System 1’ (or ‘Type 1’ processing, Evans, The aim of this study was to investigate the effects of 2008) and ‘System 2’ (or ‘Type 2’ processing, Evans, 2003; hydration status, drinking water, and thirst on a range of cog- Sloman, 1996). nitive processes. Based on the findings of previous studies To be able to measure what happens when Type 1 pro- (Benton & Burgess, 2009; Edmonds & Jeffes, 2009, Booth cessing is in conflict with the more reflective Type 2 pro- et al., 2012, Edmonds et al. 2013, Rogers et al., 2001, Benton cessing, Frederick (2005) developed the CRT. This brief et al., 2016), we expected performance for sustained atten- test consists of short maths puzzles, which can appear at tion and executive function tests to be positively affected first glance as having a very obvious answer. However, the by water consumption (Hypothesis 1). It was predicted that respondent’s initial intuitive response is incorrect, and they since water consumption has positive effects on typical cog- can only arrive at the correct response if they suppress their nitive processing tests, this should extend to judgment and initial heuristic answer and engage in more reflective think - decision-making tasks (Hypothesis 2) and that attention and ing. The CRT has been linked to performance on numerous executive performance scores (here: set shifting) correlate judgment and decision-making tasks, such as risk taking, with judgement and decision-making tasks (Hypothesis 3). temporal discounting, and use of heuristic thinking (Fred- If so, judgement and decision-making scores may be dif- erick, 2005; Toplak, Stanovich, & West, 2011). ferentially associated—depending on water supplementa- Hydration has been previously linked to activation in tion—with scores of tasks measuring general levels of atten- the anterior cingulate cortex (ACC), and the latter in turn tion (Hypothesis 4a), mental manipulation of information has been proposed to be an important component of frontal (Hypothesis 4b) or inhibition performance (Hypothesis 4c). attentional control systems (Braver, Barch, Gray, Molfese, The last two predictions are mainly based on the dual pro- & Snyder, 2001; Bush et al., 1998; Pardo, Pardo, Janer, & cess theories predicting monitoring of type 1 processing by a 1 3 Psychological Research reflective system that either inhibits automatic thinking pro- et al., 2013). This was converted to a percentage where a cesses (Frederick, 2005; Kahneman, 2011), increases think- higher percentage indicated a higher level of thirst. ing performance in terms of mental manipulation abilities, or both (Toplak et al., 2011). Cambridge neuropsychological test automated battery Importantly, we controlled for whether these effects were (CANTAB) moderated by thirst or physiological dehydration status. Uri- nary osmolality analysis was used to assess hydration status The CANTAB eclipse software (Sahakian & Owen, 1992) to formally examine a potential association between hydra- contains an array of tests used to assess cognitive perfor- tion status, thirst, drinking water and cognitive performance. mance. We administered six tests from this platform: the motor screening test (MOT), the simple reaction time (SRT), the choice reaction time (ChoiceRT), the big/little circle (BLC), the intra-extra dimensional set shifting (IED), and Methods the rapid visual information processing (RVP). The IED, RVP and ChoiceRT assess executive functions including Participants visual attention, which is the focus of this report. ChoiceRT is a 2-choice reaction time test with stimulus 31 participants (14 males) were recruited for this study and response uncertainty introduced by having two possi- through advertisements placed at the University of East ble stimuli and two possible responses. Participants were London (UEL) and on psychology websites, via emails instructed to press the left-hand button if the stimulus (an to UEL students, and through friends. A pre-participation arrow) was displayed on the left-hand side of the screen, health questionnaire was sent to interested individuals, to and the right-hand button if the stimulus was displayed on exclude persons for whom overnight fasting might have been the right-hand side of the screen. A practice stage (24 trials) a potential health risk (e.g., pregnant women and people was followed by two assessment stages (50 trials each). The suffering from diabetes or a heart condition). The ages of dependent variable was reaction time. participants ranged from 21 to 46 years (mean 31 years; IED is a test of rule acquisition and reversal. Two pat- SD 7.24). terns are displayed on the screen, first simple (colour-filled shapes) and then compound (white lines overlying colour- Tasks and questionnaires filled shapes). The participant must learn which of the two stimuli was correct by trial and error learning. When six Task were presented to participants in the following order. consecutive correct responses were recorded, the contingen- Parallel forms of all tasks were administered counterbal- cies were reversed and this pattern of stimulus addition and anced across water condition and task version. reversal continued for nine blocks. If the participant failed to reach six consecutive responses after 50 trials, the test was terminated. The dependent variable for this task was the total International positive and negative affect schedule errors committed. short form (I-PANAS-SF) RVP is a sensitive measure of general performance and in particular of visual sustained attention. Numbers appear The I-PANAS-SF scale is a shorter version of the original one at a time in a box in the centre of the screen at the rate of PANAS consisting of 10 items instead of 20 (Thompson, 100 digits per minute. Participants were instructed to press 2007) used to measure general affect. Half of the emotion the button on the press pad whenever they spotted a target words presented reflect negative affect states (ashamed, sequence of three consecutive numbers. A practice stage afraid, hostile, nervous, upset) and the other half reflect (lasting 2 min) in which participants were prompted as to positive affect states (active, alert, attentive, determined, when a sequence had begun and when to press the button inspired). Participants rated their positive and negative affect was followed by a test stage (lasting 4 min) in which no cues on a 5-point scale that ranged from “very slightly or not at were displayed and the participant had to spot three different all” (1) to “extremely” (5). sequences on their own. Target sequences occurred at the rate of 16 every 2 min. The measured dependent variable Thirst scale was total error rate. Participants were asked to indicate their level of thirst by Measuring cognitive reflection performance marking an X on a continuous horizontal line (17.8 cm) with anchors indicating “not at all” to “very thirsty” (Edmonds To assess judgement and decision-making performance, we employed tasks that are typically used to assess the use 1 3 Psychological Research of heuristic (automatic) processing that can be overcome long would it take for the patch to cover half of the lake? by reflective (controlled and analytic) thinking (follow - (Intuitive answer: 24; correct answer: 47). ing largely Toplak et al., 2011). This consisted of nine vignettes or puzzles in total per session. Six of these were Measurement of hydration status heuristics-and-biases vignettes from widely cited publica- tions that reflect important aspects of rational thought such A Vitech Advanced Multi Sample Micro freezing point as probabilistic reasoning, hypothetical thought, theory osmometre from Advanced Instruments Inc. was used to justification, scientific reasoning, and the tendency to think determine urine osmolality (mOsm/kg) to assess partici- statistically. Each answer to a heuristic vignette task was pants’ hydration status. A higher value indicates a greater scored as correct or incorrect (1 or 0 score), resulting in a degree of dehydration. According to the US National Insti- total maximum score of 6 (per session). The battery was tutes of Health, a concentration of 500–800 mOsm/kg is comprised of the following: considered normal, whereas a 12–14 h fluid restriction should yield a value in excess of 850 mOsm/kg (Chern- 1. Causal base rate (Fong, Krantz, & Nisbett, 1986). ecky & Berger, 2012). A higher value indicates a greater 2. Sample size (Tversky & Kahneman, 1974). degree of dehydration. 3. Gambler’s fallacy (Toplak et al., 2011). 4. Conjunction fallacy (Tversky & Kahneman, 1983). 5. Bayesian reasoning (Doherty & Mynatt, 1990). Procedure 6. Sunk cost (Arkes & Blumer, 1985). A pre-participation health questionnaire was sent to inter- Example of sample size: ested individuals, to exclude persons for whom overnight A certain town is served by two hospitals. In the larger fasting might have been a potential health risk (e.g., preg- hospital about 45 babies are born each day, and in the nant women and people suffering from diabetes or a heart smaller hospital about 15 babies are born each day. As you condition). They were also provided with an empty sample know, about 50% of all babies are boys. However, the exact container in which they supplied their waking urine sam- percentage varies from day to day. Sometimes it may be ple, which they also brought with them to each of their higher than 50%, sometimes lower. For a period of 1 year, sessions. each hospital recorded the days on which more than 60% Participants visited UEL’s Psychology Research Suite of the babies born were boys. Which hospital do you think on two occasions, 1 week apart, after having fasted (no recorded more such days? food or drink) from 9 p.m. the night before. Participants were asked to collect a urine sample upon waking (in ster- (a) The larger hospital. ile sample pots already provided), which they brought with (b) The smaller hospital. them. Testing took place in the mornings (8 a.m.–11 a.m.). (c) About the same (that is, within 5% of each other). To standardise the water content of breakfast, before each testing session, participants received a choice of cereal bar In addition to the vignettes inducing heuristic thinking, (113 kcal or 119 kcal). On one occasion (counterbalanced the Cognitive Reflection Test (CRT; Frederick, 2005) was across participants), they were also given a 500 ml bottle used. The CRT is designed to measure participants’ ten- of water (at room temperature). Participants were explic- dency to override an intuitive first response and to engage itly and clearly instructed to drink as much as they wanted in reflective thinking to arrive at the correct answer (simi- before beginning the tasks. There was no time pressure, lar to the mechanism proposed to work in solving heuris- but all participants stopped drinking after 2  min. They tic vignettes, Kahneman, 2011). The dependent variable were not allowed to continue drinking during testing. was the total number of correct responses (maximum of Participants then completed the tasks in the order they 3 per session). The original CRT comprised of only three have been described above. At the end of testing, they questions. We used the extended version by Toplak et al. were asked to provide another urine sample. The second (2014) resulting in different three questions in each of the session followed the same procedure and at the end of the two sessions. The answers to the six heuristic vignettes second session they were debriefed and compensated for and the three CRT puzzles formed the cognitive reflection their time and participation. Tasks in both sessions were score (a maximum of nine correct answers per session). completed in approximately 1 h. An example of the CRT is the following: in a lake, there is The order of water supplementation and tasks admin- a patch of lily pads. Every day, the patch doubles in size. If istered was counterbalanced so that 15 participants had it takes 48 days for the patch to cover the entire lake, how water in their first session and 14 in their second session, and 15 had version A of decision-making tasks in their 1 3 Psychological Research first session and 14 had version B of decision-making tasks Water consumption and hydration status effects in their second session. on thirst and mood scales Data analysis In the water condition, participants drank a mean of 303.44 ml (SD 158.21; range 50–500 ml). To test whether The main aim of this study was to investigate the effect of people were indeed dehydrated in the no water condition water supplementation on cognitive performance. To test after test as well as on both mornings, we ran a 2 (water vs hypothesis 1 and 2, the data was subjected to a series of no water) by 2 (waking vs end of test) ANOVA on osmolal- mixed analyses of variance (ANOVA) in which water sup- ity readings. There was no effect of day, F < 1, but a (1,28) plementation (water/no water given) was a within-partic- main effect of test time, (F = 5.96, p = .021, η = 0.176), (1,28) p ipants factor, and order (water first/no water first), thirst as well as an interaction, (F = 6.231, p = 0.019, (1,28) (thirsty/not thirsty), and urine osmolality (high/low) were η = 0.182). Whereas there was no difference between between-participants factors. The same analyses were also hydration readings in the water condition before (M = 735, performed for the combined cognitive ref lection scores. SD = 252) and after (M = 758, SD = 235) testing, there was For thirst and hydration, median splits were performed a difference in the no water day, with readings lower before grouping participants as either thirsty/not thirsty and (M = 693, SD = 218) than after (M = 813, SD = 217) testing hydrated/not hydrated based on the respective medians of (see Fig.  1). This suggests that on a group level, partici- 63% and 827.5 mOsm/kg on the ‘no water day’. The post- pants were reasonably dehydrated (osmolality readings of ca test osmolality data was used in the present analyses. The 700–800 mOsmo/kg), but also that in the no water day, the pre-test data was used to confirm fasting (see “Results”). dehydration became significantly worse during the morning To investigate hypothesis 3 and 4, correlation analy- compared to the water day (Edmonds et al., 2013). Thus, ses were also performed in an attempt to tease apart a water supplementation on the water day prevented further possible relationship between performance on the judg- dehydration, which seemed to happen on the no water day ment and decision-making tasks and performance on the as testing went on through the morning. Thirst ratings also ChoiceRT, IED, and RVP. confirmed that participants arrived thirsty: participants rated themselves as having greater subjective thirst on the occasion that they were not offered water (F = 46.112, (1,27) p < 0.001). Results The responses to the I-PANAS-SF mood scale were mostly unaffected by water supplementation, thirst, order Data from two participants could not be analysed because and osmolality. There were two exceptions to this state- they did not return for the second session. The final sam- ment: there was a water supplementation x order interac- ple size was 29 participants (16 females). tion for “attentive” and an osmolality effect on “inspired”. Fig. 1 Mean osmolality readings for participants on different sessions and time points during testing days (before—“waking”—and after tests). Error bars are standard errors of the mean. Significant differences (using “asterisk” to denote p > .05) between condi- tions are indicated 1 3 Psychological Research Participants who received water in their first session factor), again found similar effects for the independent vari - reported being more “attentive” on that occasion compared ables on CANTAB scores, all Fs < 1 (except the trend of to their second session in which they did not have any water water for ChoiceRT, with p values between 0.065 and 0.071. (F = 16.00, p < 0.001). In the case of urine osmolal- (1,27) ity, dehydrated participants (as evidenced by higher urine Water consumption effects on judgment osmolality) rated themselves as significantly less “inspired” and decision‑making performance (F = 4.276, p = 0.048). There was no effect of thirst (1,27) (high vs low scorers) on any of the items presented in the Three mixed-design ANOVAs were performed with total I-PANAS-SF mood scale. correct score for the combined judgement and decision-mak- ing tasks (six heuristic vignettes and three CRT vignettes in Water consumption effects on executive functions each session, see Methods) as the dependent variable, analo- gous to the ANOVAs for the executive function tests above. Mean scores on CANTAB tests were screened for normal The within-participant factor in each ANOVA was water distribution and outliers, using the interquartile range rule of supplementation (water vs no water). The between factors g = 3 (Hoaglin et al, 1986). Only one RVP errors data point in the respective ANOVAs were order (water first session or was substituted with RVP misses in one condition for one water second session), hydration (osmolality: high or low; participant who had a very high RVP false alarm rate in one i.e., dehydrated or hydrated), and finally thirst (high or low condition. For all other participants, the RVP total errors after median split). In all three ANOVAs, there was a main were calculated as the sum of the number of false alarms effect of water supplementation (Table  4), for the ANOVA and number of misses. on water and order (F = 7.37, p = 0.011, η = 0.215), (1,27) p Performance on each of the CANTAB tasks was analysed water and hydration (F = 7.44, p = 0.013, η = 0.209), (1,27) p using mixed-design ANOVAs, one separately for effects of and water and thirst (F = 7.69 p = 0.012, η = 0.212). (1,27) p order, thirst, and osmolality. The within-participants factor Participants scored overall higher on the judgment and deci- in each ANOVA was water supplementation (water vs no sion-making tasks in conditions in which they received water water). The between factors in the respective ANOVAs were compared to the no water day (Fig. 2). There were no simple order (water r fi st session or water second session), hydration main effects from factors order, F 2.730, p = .110, hydra- (1,27) [osmolality: high > 827.5 mOsm/kg or low < 827.5 mOsm/ tion F < 1, or thirst, F = 2.33, p = .138. There were (1,27) (1,27) kg; i.e., dehydrated (15) or hydrated (14)], and finally thirst also no interaction effects involving order, all F s < 1. Water (high or low after median split; 14 participants classified supplementation therefore had a positive effect on scores as thirsty and 15 as not thirsty). There were no significant across the battery of judgment and decision-making tasks, effects or trends for the factor order or water (see Tables  1, 2, relatively independent of levels of thirst and hydration (on 3), bar two exceptions. There were trends for ChoiceRTs to the no water day), or order. The ANCOVAs using thirst and be generally faster in the water conditions (p values between osmolality (mean-centred) as co-variates instead of median 0.066 and 0.073; Tables 1, 2, 3), and there was a significant splitting found the same patterns effects on cognitive reflec- interaction for water and order in the RVP tasks (Table 1), tion scores, with no main or interaction effect of the co- with more errors in the water condition (M = 17.80, variates (all Fs < 1). SD = 4.72) compared to the no water condition (M = 21.13, This result confirmed hypothesis 2, that water supplemen- SD = 5.55) when participants received water in their first ses- tation increased cognitive reflection scores, and this result sion, p = 0.007, but vice versa when they received it second, was not qualified by any interaction. p = 0.057 (water: M = 22.07, SD = 3.15; no water: M = 20.50, SD = 3.65). Therefore, hypothesis 1 could not be retained. Correlation analysis Regarding the main effects of between-subjects vari- ables, there was a marginal effect of order on the ChoiceRT, To investigate the possible relationship between hydration F = 4.034, p = .055, with higher RTs in the first session variables, judgment scores and executive functions for dif- (1,27) (M = 327, SD = 12) than in the second (M = 290, SD = 13). ferent water supplementation conditions, we performed cor- Otherwise, there were no effects, all Fs < 1, except for IED relation analyses. Table 5 shows differing degrees of associa- errors and order, F = 1.503, p = .23, and order effects on tions depending on whether the data used was taken from the (1,27) RVP error rates, F = 2.161, p = .153. Controlling for the day participants received water or not. (1,27) amount of water each participant drank (using ANCOVAs) There were significant correlations between cognitive did not change the pattern of effects, all F s < 1, except reflection scores and ChoiceRT (water: r = − 0.473, p = (1,28) for a similar trend as above for ChoiceRT F = 3.53, 0.010; no water: r = − 0.579, p = 0.001) and IED errors (1,28) p = .71. Additional ANCOVAs using thirst and osmolality (water: r = − 0.533, p = 0.003; no water: r = − 0.578, as co-variates (rather than median split as a between-group p = 0.001) on both days, water and no water, respectively. 1 3 Psychological Research Fig. 2 Mean scores for the com- bined judgment and decision- making tasks comparing perfor- mance on the day participants received water with the day they did not. Water consumption has a significant effect on scores, with participants scoring better on the day they did receive water Table 1 CANTAB test means, SDs and F ratios by water condition (water/no water) and order (water first/no water first) Task Water first No water first Results from the omnibus statistical analysis; those with p < .05 in bold Water No water Water No water M SD M SD M SD M SD ChoiceRT 317.81 51.63 336.33 74.21 288.73 31.55 292.27 35.05 Water F = 3.476, p = 0.073 (1,27) Water × order F = 1.600, p = 0.217 (1,27) IED total errors 20.07 10.53 18.33 11.76 13.79 9.31 15.43 9.99 Water F = 0.002, p = 0.966 (1,27) Water × order F = 2.615, p = 0.117 (1,27) RVP errors 10.53 5.64 7.60 6.34 5.50 3.43 7.42 4.21 Water F = 0.335, p = .568 (1,26) Water × order F = 14.259, p = 0.001 (1,26) Table 2 CANTAB test means, SDs and F ratios by water condition (water/no water) and post-testing urine osmolality (low/high) as measured on the day participants did not receive any water Task Low osmolality High osmolality Results from the omnibus statistical analysis; those with p < .05 in bold Water No water Water No water M SD M SD M SD M SD ChoiceRT 304.26 46.83 319.05 75.54 303.31 44.61 311.34 48.19 Water F = 3.548, p = 0.070 (1,27) Water × osmo F = 0.312, p = 0.581 (1,24) IED total errors 17.53 11.04 18.67 11.32 16.50 9.80 15.07 10.41 Water F = 0.019, p = 0.891 (1,27) Water × osmo F = 1.446, p = 0.240 (1,24) RVP errors 8.57 5.71 7.64 5.51 7.71 5.21 7.50 5.53 Water F = 0.552, p = 0.464 (1,26) Water x Osmo F = 0.169, p = 0.684 (1,26) There was also a significant correlation between CRT executive function tasks being associated with higher cogni- scores and RVP errors on the water day only (r = − 0.451, tive reflection performance. Hypothesis 3 was therefore con- p = 0.014). All correlations were in the predicted direction firmed—performance on executive function tasks (though with better performance (lower errors or shorter RTs) in only in the water condition for RVP) was associated with 1 3 Psychological Research Table 3 CANTAB test means, SDs and F ratios by water condition (water/no water) and thirst (low/high) as measured on the day participants did not receive any water Task Low thirst High thirst Results from the omnibus statistical analysis; those with p < .05 in bold Water No water Water No water M SD M SD M SD M SD ChoiceRT 303.96 41.45 309.75 38.81 303.57 49.87 320.75 80.94 Water F = 3.676, p = 0.066 (1,27) Water × thirst F = 0.903, p = 0.350 (1,24) IED total errors 16.29 10.61 15.43 10.17 17.73 10.31 18.33 11.62 Water F = 0.011, p = 0.916 (1,27) Water × thirst F = 0.124, p = 0.728 (1,27) RVP errors 8.14 5.75 7.64 4.73 8.14 5.20 7.50 6.21 Water F = 0.536, p = 0.470 (1,27) Water × thirst F = 0.024, p = 0.879 (1,27) Table 4 Cognitive reflection Water No water Results from the statistical analysis score means, SDs and F ratios by water condition (water/ M SD M SD no water) and post-testing Water first 4.80 2.18 4.27 1.62 Water F = 7.374, p = 0.011, η = 0.215 urine osmolality (low/high) (1,27) p as measured on the day Water second 6.00 1.47 5.07 1.77 Water × order F = 0.539, p = 0.469 (1,27) participants did not receive any 2 Osmo low 5.47 2.44 4.53 1.92 Water F = 7.439, p = 0.013, η = 0.209 (1,27) p water Osmo high 5.29 1.27 4.79 1.53 Water × osmo F = 0.651, p = 0.427 (1,27) Thirst low 4.87 2.13 4.27 1.53 Water F = 7.688, p = 0.012, η = 0.212 (1,27) p Thirst high 5.93 1.59 5.07 1.86 Water × thirst, F = 0.226, p = 0.639 (1,27) Table 5 Correlations between cognitive reflection performance vignettes. Some approaches in the dual systems framework scores and water consumption, urine osmolality, thirst, CANTAB (e.g., Evans & Stanovich, 2013) further implicate mental tasks and for both days (participants received/did not receive water) simulation performance, the ability to maintain and sym- (N = 29) bolically manipulate separate mental representations of a Measured variable Water No water problem. ChoiceRT latency difference scores were log-transformed Water consumed (ml) 0.208 – to reduce potential issues of positive skew and normality of Urine osmolality (post) 0.132 0.107 residuals. Results of the multiple linear regression indicated Thirst −0.168 0.169 that there was a combined significant effect of differences in ChoiceRT (RT) −0.473** −0.579*** ChoiceRT and IED (errors) explaining differences in cogni- IED (errors) −0.533** −0.578** tive reflection scores, (F = 3.765, p = 0.037, R = 0.224). (2,26) RVP (errors) −0.451* −0.148 ChoiceRT difference (t = − 2.244, p = 0.034) was a signifi- *p ≤ 0.05. **p < .01. ***p < .001 cant predictor in the model, but not IED error difference (t = − 1.543, p = 0.135) (Table 6). Adding RVP errors (difference higher performance in the judgment and decision-making scores) as a predictor variable again showed a relationship tasks. for cognitive reflection scores with ChoiceRT (t = − 2.343, Finally, a linear regression analysis was performed p = 0.027) but not RVP errors (t = − 1.005, p = 0.324), with using difference scores (water–no water). The difference the overall model marginally significant, (F = 4.899, p = (2,25) in cognitive reflection scores between the no water and the 0.058, R = 0.254). Thus, hypothesis 4c was retained: cogni- water condition served as the dependent variable (crite- tive reflection scores for sessions in which water was given rion) and the difference (between water and no water day) were differentially influenced by ChoiceRT scores compared in ChoiceRT and IED errors as independent variables. The to sessions in which water was not given—the higher the regression model tested whether differences (between ses- differences in ChoiceRT latencies (and therefore the worse sions) in the executive function tasks were associated with the inhibition performance between no water and water con- differences in the cognitive reflection scores. Recall that dition), the lower the improvement of cognitive reflection the hypothesis was based on the premise by dual process scores from no water to water condition. theories that increased inhibition processes are related to Hypothesis 4a and 4b were therefore not retained—dif- increased performance in CRT-like puzzles and heuristic ferences in tasks measuring attention performance (RVP) or 1 3 Psychological Research Table 6 Summary of regression analysis for variables predicting cognitive reflection score differences (N = 29) between sessions Source B SE B β t p LBCI 95% UBCI 95% ChoiceRT— −4.357 1.942 −0.388 −2.244 0.034* −8.349 −0.366 diff IED (errors)— −0.066 0.043 −0.267 −1.543 0.135 −0.155 0.022 diff mental simulation (IED) between the water supplementation (Edmonds et al., 2013). However, we have failed to replicate conditions were not associated with the difference in cogni - the significant findings pertaining to participants’ subjective tive reflection scores. ratings of thirst as a moderator of the effects of water supple- mentation on most the measures assessed (including mood ratings). This could be an idiosyncrasy of the particular sam- Discussion ple population or it could indicate individual differences in feelings of subjective thirst. For example, several partici- The current study is to our knowledge the first to report pants (N = 7) in this study spontaneously expressed that they increased cognitive reflection performance (and, by exten - were seldom thirsty, so it was perhaps not surprising that sion, increased judgment and decision-making performance, even on the occasion when they were not given any water, Frederick, 2005; Toplak et al., 2011) after water consump- they indicated a relatively low level of thirst. Nonetheless, tion. When thirst, hydration status, and mood state were con- as elucidated by the relevant statistical analysis on a group trolled for, water supplementation increased performance on level, participants did report experiencing greater levels of an overall composite score from widely used judgment and subjective thirst on the occasion they did not receive any decision tasks (judgment vignettes eliciting heuristic think- water. ing, simple maths puzzles requiring cognitive reflection). Our main result is the significant effect of water supple- These scores were related to inhibition processing speed and mentation on performance on the judgement and decision- executive functions (ChoiceRT and IED), but not attentional making tasks (heuristics and biases, cognitive reflection performance (RVP) or feelings of general attentiveness. The test)—participants performed better on the occasion on experimentally induced differences in judgement and deci- which they received water. This finding cannot be easily sion performances between water days and no water days dismissed as a result of demand characteristics (i.e., simply were associated with differences in Stroop-like task perfor - being given a drink increasing motivation, or expectation of mances (and Simon task) generally associated with inhibi- doing better), because we did not find an influence of water tion processes. Before we turn to these effects in detail, we supplementation on mood effects (see also Edmonds et al., discuss the physiological factors that could have influenced 2013 who show that expecting water supplementation does this result. not explain increased performance in attention tasks) nor on In general, there were no effects of water, thirst or hydra- other cognitive tasks (IED, RVP). Therefore, we interpret the tion status (except for PANAS ‘attentive’ scores, but those effects on cognitive reflection scores as substantially driven were not associated with cognitive reflection performance) by water supplementation. on the measures of mood used in this study. Some stud- The tasks used here were aimed at assessing ‘slow’ pro- ies have previously reported links between dehydration cessing (reflective thinking) vs ‘fast’ processing (heuristic and mood ratings (Shirreffs et al., 2004), and water supple- thinking; Kahneman, 2003), with the particular aim to inves- mentation and mood ratings (Edmonds et al., 2013), while tigate potential processes that override decisions reached by others report that water supplementation does not affect Type 1 (De Neys & Glumicic, 2008). Inhibition performance mood (Edmonds et al., 2013). Furthermore, it may be that has been shown to be influenced by water supplementation whether mood affects dehydration may depend on the man- in previous research (Edmonds et al., 2013) and could thus ner in which dehydration is achieved: Shirreffs et al. (2004) modulate the effect of water supplementation on cognitive induced dehydration by fluid restriction, whereas Edmonds reflection performance. Indeed, performance on the Stroop- et  al. (2013) reported effects on water supplementation. like ChoiceRT task correlated with judgment and decision At any rate, the main finding here is that mood (and hence performance score in both conditions, as was predicted expectation effects) does not explain the findings for the (and replicated previous results, e.g., Toplak et al., 2011). effect of water on cognitive reflection tasks. In addition, regression analysis suggests a link between inhi- Previous studies have revealed that both water sup- bition performance (as measured by the ChoiceRT) and the plementation and thirst impact on cognitive performance effect of water supplementation on decision performance: 1 3 Psychological Research As ChoiceRT performance is affected by supplementation 2000). Similarly, Evans and Stanovich (2013) propose that (although effect sizes are small), so is the performance on the reflective system requires executive processes beyond the heuristic vignettes and puzzles. Of course, we cannot inhibition to enable ‘cognitive decoupling’, that is the directly infer causation, but it is noteworthy that most dual ability for mental simulation and abstract thinking. It is process theories of thinking and deciding (De Neys & Glu- this ability that potentially allows the independent mental micic, 2008; Evans & Stanovich, 2013; Kahneman, 2011) representation of information in math-like puzzles (CRT) predict that cognitive reflection performance relies on moni- and vignettes (heuristics) shown to participants. Indeed, toring and consequently inhibiting the pre-potent responses our findings of strong correlations between IED (set shift- related to heuristic thinking. It is the successful monitoring ing) and cognitive reflection tasks strongly indicates that and inhibition that consequently decreases biased judgments. some form of cognitive decoupling underlies Type 2-like Our finding that particular executive processes correlate processing. But here again, there was no indication that with cognitive reflection tasks is also roughly in line with IED—as a proxy measure for mental simulation- modu- the psychobiological literature, especially the notion of a lates the relationship between water supplementation and possible role of a behavioural inhibition in judgement per- cognitive reflection scores, unlike what we found with the formance. For example, fMRI studies found that dehydra- ChoiceRT task. tion directly affects the blood flow to the anterior cingulate Our findings are therefore the first that show a tentative cortex (Farrell et al., 2008), which is linked to inhibition link between water supplementation (after dehydration), (e.g., Stroop) performance. When comparing incongruent inhibition performance, and judgement and decision- and control conditions, the majority of such studies report making processes. If future research confirms the effect maximal differential activation occurring in the anterior cin- of water supplementation (and a possible role of dehy- gulate cortex (Bench et al., 1993; Bush at al., 1998; Carter dration) on decision-making performance, the underlying et al., 1995; Carter et al., 2000; Derbyshire, Vogt, & Jones, cognitive–physiological mechanisms may be more com- 1998; Pardo et al., 1990). Although this area shows great- plicated. Hydration has been linked to a range of inhibi- est activation in the incongruent condition of the Stroop, tory or excitatory effects, leading to cognitive improve - the congruent condition (facilitation) has also been shown ments or impairments. For example, chronic dehydration to increase activation as compared to a control condition in animals increases the release of the neurotransmitters (Bench at al., 1993; Carter et al., 1995). Thus, even though gamma-aminobutyric acid (GABA) and glutamate, which the choice response times may be deemed a somewhat indi- have inhibitory and excitatory effects, respectively (Di & rect inhibition measure, both congruent and incongruent Tasker, 2004). Furthermore, dehydration has been shown trials (and latency data) may indicate inhibition processes. to increase the release of the stress hormone cortisol Furthermore, previous work has shown that complex pro- (Francesconi et al., 1984), and elevated cortisol levels have cessing speed measures are substantially correlated with been associated with impaired cognitive function (Green- executive control measures but not with simpler speed meas- dale et al., 2000; Kirschbaum et al., 1996). In the present ures (e.g., Cepeda et al., 2013). However, future research data, similar conflicting effects may therefore account needs to further elucidate the exact mechanisms of inhibition for the difficulty in establishing stronger links between and facilitation linked to water supplementation. Moreover, executive functions and cognitive reflection performance different tasks may tap into different inhibition processes. in different water supplementation (and hence hydration) For example, Khng and Lee (2014) found performance on conditions. the Stroop tasks largely independent from performance on a In conclusion, we find a clear effect of water supple- Stop-signal task, indicating potentially different underlying mentation (after dehydration) on decision-making per- inhibition processes. Although the ChoiceRT task employed formance when thirst is controlled for. The challenge here contains conditions that require Stroop-like inhibition for future studies will be to further clarify the relation- processes, other tasks may help establish better models ship between physiological and cognitive mechanisms. explaining the relationship between inhibition, executive Researchers will need to employ executive function tests functions, and cognitive reflection performance. In any that are sensitive to different types of inhibition mecha- case, though intriguing, the link between a task involving nisms and other executive processes, as well as measuring inhibition processes and cognitive reflection performance effects stemming from physiological hydration and thirst. found here needs to be interpreted with caution until further Acknowledgements We thank Dr. Caroline Edmonds for her com- replicated with other measures of inhibition. ments on the manuscript. We would like to thank Dr. Paula Booth for In addition to inhibition, Toplak et al. (2011) found that her help and support in running this study, and for her guidance with the cognitive reflection task (CRT) also correlates with regard to the urine osmolality analysis in particular. measures of cognitive ability (see also Stanovich & West, 1 3 Psychological Research Funding This work was supported by a student grant to Olivia Patsalos https ://doi.or g/10.1002/(SICI)1097-0193(1998)6:4%3C270 from the European Hydration Institute. ::AID-HBM6%3E3.0.CO;2-0 Carter, C. S., Macdonald, A. M., Botvinick, M., Ross, L. L., Stenger, V. A., Noll, D., & Cohen, J. D. (2000). Parsing executive processes: Compliance with ethical standards Strategic vs. evaluative functions of the anterior cingulate cortex. Proceedings of the National Academy of Sciences of the United Conflict of interest Dr. Volker Thoma declares that he has no conflict States of America, 97(4), 1944–1948. https ://doi.org/10.1073/ of interest. Olivia Patsalos declares that she has no conflict of interest. PNAS.97.4.1944. Carter, C. S., Mintun, M., & Cohen, J. D. (1995). Interference and Ethical approval All procedures performed in studies involving human facilitation effects during selective attention: An H215O PET participants were in accordance with the ethical standards of the insti- study of Stroop task performance. NeuroImage, 2(4), 264–272. tutional and/or national research committee and with the 1964 Helsinki https ://doi.org/10.1006/nimg.1995.1034. declaration and its later amendments or comparable ethical standards. Cepeda, N. J., Blackwell, K. A., & Munakata, Y. (2013). Speed isn’t everything: Complex processing speed measures mask individ- Informed consent Informed consent was obtained from all individual ual differences and developmental changes in executive control. participants included in the study. Developmental Science, 16(2), 269–286. Chernecky, C., & Berger, B. (2012). Laboratory tests and diagnostic procedures. 6th edn. Saunders. eBook ISBN: 9781455745029. Open Access This article is distributed under the terms of the Crea- Cokely, E. T., & Kelley, C. M. (2009). Cognitive abilities and supe- tive Commons Attribution 4.0 International License (http://creat iveco rior decision making under risk: A protocol analysis and process mmons.or g/licenses/b y/4.0/), which permits unrestricted use, distribu- model evaluation. Judgment and Decision Making, 4(1), 20–33. tion, and reproduction in any medium, provided you give appropriate https ://doi.org/10.1016/j.jbank fin.2009.04.001. credit to the original author(s) and the source, provide a link to the De Neys, W., & Glumicic, T. (2008). Conflict monitoring in dual pro- Creative Commons license, and indicate if changes were made. cess theories of thinking. Cognition, 106(3), 1248–1299. https :// doi.org/10.1016/j.cogni tion.2007.06.002. Derbyshire, S. W. G., Vogt, B. A., & Jones, A. K. P. (1998). Pain and Stroop interference tasks activate separate processing modules in References anterior cingulate cortex. Experimental Brain Research, 118(1), 52–60. https ://doi.org/10.1007/s0022 10050 254. Arkes, H. R., & Blumer, C. (1985). The psychology of sunk cost. Di, S., & Tasker, J. G. (2004). Dehydration-induced synaptic plasticity Organizational Behavior and Human Decision Processes, 35(1), in magnocellular neurons of the hypothalamic supraoptic nucleus. 124–140. https ://doi.org/10.1016/0749-5978(85)90049 -4. Endocrinology, 145(11), 5141–5149. h t tp s : // d o i. o r g /1 0 .1 2 10 / Bar-David, Y., Urkin, J., & Kozminsky, E. (2005). The effect of vol- en.2004-0702. untary dehydration on cognitive functions of elementary school Document_not_found. (2018). Document not found (Frederick, 2005). children. Acta Paediatrica, 94(11), 1667–1673. h t t p s : / / d o i . Doherty, M. E., & Mynatt, C. R. (1990). Inattention to P(H) and to org/10.1080/08035 25050 02546 70. P(D/~H): A converging operation. Acta Psychologica, 75(1), Bench, C. J., Frith, C. D., Grasby, P. M., Friston, K. J., Paulesu, E., 1–11. https ://doi.org/10.1016/0001-6918(90)90063 -L. Frackowiak, R. S., & Dolan, R. J. (1993). Investigations of the Edmonds, C. J., & Burford, D. (2009). Should children drink more functional anatomy of attention using the Stroop test. Neuropsy- water? The effects of drinking water on cognition in children. chologia, 31(9), 907–922. Appetite, 52(3), 776–779. https ://doi.or g/10.1016/j.appe t Benton, D. (2011). Dehydration influences mood and cognition: .2009.02.010. A plausible hypothesis? Nutrients, 3, 555–573. https ://doi. Edmonds, C. J., Crombie, R., Ballieux, H., Gardner, M. R., & Dawkins, org/10.3390/nu305 0555. L. (2013). Water consumption, not expectancies about water con- Benton, D., & Burgess, N. (2009). The effect of the consumption of sumption, ae ff cts cognitive performance in adults. Appetite, 60(1), water on the memory and attention of children. Appetite, 53(1), 148–153. https ://doi.org/10.1016/j.appet .2012.10.016. 143–146. https ://doi.org/10.1016/j.appet .2009.05.006. Edmonds, C. J., & Jeffes, B. (2009). Does having a drink help you Benton, D., Jenkins, K. T., Watkins, H. T., & Young, H. A. (2016). think? 6–7-year-old children show improvements in cognitive per- Minor degree of hypohydration adversely influences cognition: A formance from baseline to test after having a drink of water. Appe- mediator analysis. American Journal of Clinical Nutrition, 104(3), tite, 53(3), 469–472. https://doi.or g/10.1016/j.appet.2009.10.002 . 603–612. https ://doi.org/10.3945/ajcn.116.13260 5. Evans, J. S. B. T. (2003). In two minds: Dual-process accounts of rea- Booth, P., Taylor, B., & Edmonds, C. (2012). Water supplementation soning. Trends in Cognitive Sciences, 7(10), 454–459. https://doi. improves visual attention and fine motor skills in school children. org/10.1016/j.tics.2003.08.012. Education and Health, 30(3), 75–79. Evans, J. S. B. T. (2008). Dual-processing accounts of reasoning, judg- Braver, T. S., Barch, D. M., Gray, J. R., Molfese, D. L., & Snyder, A. ment, and social cognition. Annual Review of Psychology, 59(1), (2001). Anterior cingulate cortex and response conflict: Effects of 255–278. https://doi.or g/10.1146/annure v.psych.59.10300 6.09362 frequency, inhibition and errors. Cerebral Cortex, 11(9), 825–836. https ://doi.org/10.1093/cerco r/11.9.825. Evans, J. S. B. T., & Stanovich, K. E. (2013). Dual-process theories Buckner, R. L., & Carroll, D. C. (2007). Self-projection and the of higher cognition: Advancing the debate. Perspectives on Psy- brain. Trends in Cognitive Sciences, 11(2), 49–57. https ://doi. chological Science, 8(3), 223–241. https://doi.or g/10.1177/17456 org/10.1016/j.tics.2006.11.004. 91612 46068 5. Bush, G., Whalen, P. J., Rosen, B. R., Jenike, M. A., McInerney, S. Fadda, R., Rapinett, G., Grathwohl, D., Parisi, M., Fanari, R., Calò, C. C., & Rauch, S. L. (1998). The counting Stroop: An interference M., & Schmitt, J. (2012). Effects of drinking supplementary water task specialized for functional neuroimaging–validation study at school on cognitive performance in children. Appetite, 59(3), with functional MRI. Human Brain Mapping, 6(4), 270–282. 730–737. https ://doi.org/10.1016/j.appet .2012.07.005. 1 3 Psychological Research Farrell, M. J., Zamarripa, F., Shade, R., Phillips, P. A., McKinley, M., Pardo, J., Pardo, P. J., Janer, K. W., & Raichle, M. E. (1990). The ante- Fox, P. T., … Egan, G. F. (2008). Effect of aging on regional rior cingulate cortex mediates processing selection in the Stroop cerebral blood flow responses associated with osmotic thirst and attentional conflict paradigm. Proceedings of the National Acad- its satiation by water drinking: A PET study. Proceedings of the emy of Sciences, 87(1), 256-259. National Academy of Sciences of the United States of America, Rogers, P. J., Kainth, A., & Smit, H. J. (2001). A drink of water can 105(1), 382–387. https ://doi.org/10.1073/pnas.07105 72105 . improve or impair mental performance depending on small dif- Fong, G. T., Krantz, D. H., & Nisbett, R. E. (1986). The effects of ferences in thirst. Appetite, 36(1), 57–58. statistical training on thinking about everyday problems. Cogni- Sahakian, B. J., & Owen, aM. (1992). Computerized assessment in neu- tive Psychology, 18(3), 253–292. https ://doi.org/10.1016/0010- ropsychiatry using CANTAB: discussion paper. The Royal Society 0285(86)90001 -0. of Medicine, 85(July), 399–402. https ://doi.org/10.1177/01410 Francesconi, R. P., Sawka, M. N., & Pandolf, K. B. (1984). Hypohydra-76892 08500 711. tion and acclimation: effects on hormone responses to exercise/ Sharma, V. M., Sridharan, K., Pichan, G., & Panwar, M. R. (1986). heat stress. Aviat Space Environ Med, 55, 365– 369. Influence of heat-stress induced dehydration on mental functions. Frederick, S. (2005). Cognitive {reflection} and {decision} {making}. Ergonomics, 29(6), 791–799. Journal of Economic Perspectives, 19(4), 24–42. Shirreffs, S. M., Merson, S. J., Fraser, S. M., & Archer, D. T. (2004). Gigerenzer, G., & Goldstein, D. G. (1996). Reasoning the fast and The effects of fluid restriction on hydration status and subjective frugal way: Models of bounded rationality. Psychological Review, feelings in man. British Journal of Nutrition, 91(06), 951. 103(4), 650–669. https://doi.or g/10.1093/acprof:oso/97801 99744 Sloman, S. A. (1996). The empirical case for two systems of rea- 282.003.0002. soning. Psychological Bulletin, 119(1), 3–22. h t t p s : / / d o i . Gopinathan, P. M., Pichan, G., & Sharma, V. M. (1988). Role of dehy- org/10.1037/0033-2909.119.1.3. dration in heat stress-induced variations in mental performance. Stanovich, K. E., & West, R. F. (2000). Individual differences in rea- Archives of Environmental Health, 43(1), 15–17. https ://doi. soning: implications for the rationality debate? The Behavioral org/10.1017/CBO97 81107 41532 4.004. and Brain Sciences, 23(5), 645–726. Greendale, G. A., Kritz-Silverstein, D., Seeman, T., Barrett-Connor, Toplak, M. E., West, R. F., & Stanovich, K. E. (2011). The cognitive E. (2000). Higher basal cortisol predicts verbal memory loss in reflection test as a predictor of performance on heuristics-and- postmenopausal women: rancho bernardo study. Journal of the biases tasks. Memory & Cognition, 39(7), 1275–1289. https://doi. American Geriatrics Society, 48(12), 1655–1658.org/10.3758/s1342 1-011-0104-1. Kahneman, D. (2003). A perspective on judgment and choice: Mapping Toplak, M. E., West, R. F., & Stanovich, K. E. (2014). Assessing bounded rationality. The American Psychologist, 58(9), 697–720. miserly information processing: An expansion of the cognitive https ://doi.org/10.1037/0003-066X.58.9.697. reflection test. Thinking & Reasoning, 20(2), 147–168. https :// Kahneman, D. (2011). Thinking, fast and slow. Toronto: Doubleday doi.org/10.1080/13546 783.2013.84472 9. Canada. Tversky, A., & Kahneman, D. (1974). Judgment under uncertainty: Khng, K. H., & Lee, K. (2014). The relationship between stroop and Heuristics and biases. Science, 185(4157), 1124–1131. https :// stop-signal measures of inhibition in adolescents: Inu fl ences from doi.org/10.1126/scien ce.185.4157.1124. variations in context and measure estimation. PLoS One, 9(7), Tversky, A., & Kahneman, D. (1983). Extensional versus intui- e101356. https ://doi.org/10.1371/journ al.pone.01013 56. tive reasoning: The conjunction fallacy in probability judg- Kirschbaum, C., Wolf, O. T., May, M., Wippich, W., & Hellham- ment. Psychological Review, 90(4), 293–315. https ://doi. mer, D.. H (1996). Stress- and treatment-induced elevations of org/10.1037/0033-295X.90.4.293. cortisol levels associated with impaired declarative memory in healthy adults. Life Sciences, 58(17), 1475–1483. https ://doi. Publisher’s Note Springer Nature remains neutral with regard to org/10.1016/0024-3205(96)00118 -X. jurisdictional claims in published maps and institutional affiliations. Miyake, A., Friedman, N. P., Emerson, M. J., Witzki, A. H., Howerter, A., & Wager, T. D. (2000). The unity and diversity of execu- tive functions and their contributions to complex “Frontal Lobe” Tasks: A latent variable analysis. Cognitive Psychology, 41(1), 49–100. https ://doi.org/10.1006/cogp.1999.0734. 1 3 http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Psychological Research Springer Journals

Water supplementation after dehydration improves judgment and decision-making performance

Psychological Research , Volume OnlineFirst – Jan 21, 2019

Loading next page...
 
/lp/springer-journals/water-supplementation-after-dehydration-improves-judgment-and-decision-rSqFxyhCN3

References (50)

Publisher
Springer Journals
Copyright
Copyright © 2019 by The Author(s)
Subject
Psychology; Psychology Research
ISSN
0340-0727
eISSN
1430-2772
DOI
10.1007/s00426-018-1136-y
Publisher site
See Article on Publisher Site

Abstract

Previous research has shown that dehydration and water supplementation affect mood and cognitive performance in both adults and children on a variety of tasks that assess memory, attention, executive function, and speeded responses. Given the varied effects of water on cognition, this study explored potential effects of water supplementation, hydration status, and thirst on thinking and decision-making tasks. 29 adult participants undertook a battery of cognitive tests on two separate occasions after having fasted from the previous night. On one occasion, they were offered 500 ml of water to drink prior to testing. Measures of urine osmolality confirmed the group-level effectiveness of the dehydration manipulation. Water sup- plementation was found to improve performance on tasks measuring cognitive reflection in judgement and decision-making. This increase in performance was associated with differences in tasks implicated in inhibition processes. Drinking water after a 12-h dehydration period increased performance in judgement and decision-making tasks, and this was not explained by differences in subjective thirst or attentiveness. Introduction negatively affected (Bar-David, Urkin, & Kozminsky, 2005; Fadda et al., 2012). Several studies have found that dehydration affects not only In addition to dehydration studies, the effects of water mood but cognitive performance in both adults and children consumption on cognitive performance have been exam- (see Benton, 2011 for review). For example, Sharma et al. ined separately. Edmonds, Crombie, Ballieux, Gardner, and (1986), who dehydrated their participants to 1%, 2% and 3% Dawkins (2013) found that water consumption resulted in loss of body weight, found that performance on a memory improved performance on a visual attention task at both test and psychomotor stylus test worsened with increasing 20 min and 40 min after water supplementation when com- levels of dehydration. Similarly, exercise-induced dehydra- pared to baseline scores. In similar studies conducted in chil- tion of 2% loss of body weight was found to negatively affect dren, water supplementation was found to improve perfor- performance on memory tasks, mathematical calculation mance on memory and attention tasks (Edmonds & Jeffes, tasks, and executive function tasks (Gopinathan, Pichan, & 2009; Edmonds & Burford, 2009) and on a letter cancella- Sharma, 1988). More recently, Benton, Jenkins, Watkins, tion task, which assesses visual attention (Booth, Taylor, & and Young (2016) reported improved memory and focussed Edmonds, 2012). attention in individuals dehydrated to 1% body loss. Further- Some researches have been conducted in which the effects more, studies investigating the effects of dehydration in chil- of thirst have also been evaluated. In one study, it was found dren have found that their performance on memory tasks is that participants self-reporting as being very thirsty showed a dose-dependent improvement on a sustained attention task when drinking water (Rogers et al., 2001). More surpris- Olivia C. Patsalos and Volker Thoma are co-first authors. ingly, though, participants who reported a low thirst level showed a dose-related impairment in performance on the * Olivia C. Patsalos same task. The role of thirst on cognitive performance was olivia.patsalos@kcl.ac.uk further evidenced in a study by Edmonds et al. (2013), in Section of Eating Disorders, Department of Psychological which it was reported that thirst facilitated performance on Medicine, King’s College London, London SE5 8AZ, UK tasks requiring controlled processing (set shifting). This University of East London, Stratford Campus, Water Lane, study also reported that thirsty participants performed London E15 4LZ, UK Vol.:(0123456789) 1 3 Psychological Research significantly better on a simple reaction time task after Raichle, 1990). We therefore employed a number of execu- drinking water and that this was moderated by participants’ tive function tasks to test possible links between hydra- subjective feelings of thirst. tion and executive functions such as inhibition, set shift- In light of these findings, it is surprising that to the best ing and updating processes (Miyake et al., 2000). Research of our knowledge, no studies have been investigating the has shown that the performance on the CRT and heuristics impact of dehydration and water supplementation on judge- thinking vignettes may also depend on executive functions, ment and decision-making, crucial tasks in everyday life. such as inhibition capability (De Neys & Glumicic, 2008). This is even more surprising, as there are clear links between Specifically, if overcoming automatic thinking patterns executive functions and judgment and decision-making per- (such as the ‘implied’ but incorrect answers in the heuristic formance (Cokely & Kelley, 2009). For example, Toplak, vignettes and the CRT) is due to inhibition (of automatic— West, and Stanovich (2011) found that the cognitive reflec- ‘intuitive’ responses), then we predicted that performance on tion test (CRT) correlates positively with working memory, an alternative choice reaction time task (ChoiceRT) would inhibition, and set-shifting performance. Hence, in the pre- affect the relationship between water supplementation and sent investigation, we have employed tasks that typically tap cognitive reflection scores. This was measured in the cur - judgement and decision-making processes, aiming to capture rent study with the ChoiceRT, which measures Stroop-like any potential effects of water supplementation and thirst on inhibition performance (we used the Cambridge Neuropsy- thinking and reasoning abilities when participants are in a chological Test Automated Battery (CANTAB) software dehydrated state. package (Sahakian & Owen, 1992) to assess cognitive People are faced daily with judgements and decisions performance, see Methods). In addition to inhibition pro- which potentially put great demands on our cognitive pro- cesses, judgement and decision-making performance may cesses. Consequently, people often rely on heuristics (mental also rely on the ability to symbolically manipulate mental shortcuts) that simplify the task at hand—even in cases in representations (see Buckner & Carroll, 2007) and deal with which the task is a relatively simple one to solve (Frederick, calculations (Toplak et al., 2011). Hence, we also employ a 2005). A sizeable body of research suggests that heuristics set-shifting task, the CANTAB Intra-Extra Dimensional Set can lead to systematic deviations from logic, resulting in Shift (IED), which measures cognitive flexibility and the predictable biases and inconsistencies (Kahneman, 2003). ability to keep mental concepts or representations separate. When faced with certain probabilistic judgment problems, Finally, solving judgement vignettes may also require sus- people tend to use heuristics instead of reflective thinking tained attention (see Booth et al., 2012) and regular updat- that resembles an ‘algorithmic’ approach—the considered ing of working memory content (Miyake et al., 2000), as use of rule-based processes that should lead to normatively measured with the CANTAB Rapid Visual Processing task correct outcomes (though see Gigerenzer & Goldstein, 1996 (RVP), in which participants have to monitor a sequence of for a different view of heuristics). This led to a distinction numbers and respond when observing a target sequence of between automatic (heuristic) and reflective (analytic) think - three consecutive numbers. ing, described as ‘System 1’ (or ‘Type 1’ processing, Evans, The aim of this study was to investigate the effects of 2008) and ‘System 2’ (or ‘Type 2’ processing, Evans, 2003; hydration status, drinking water, and thirst on a range of cog- Sloman, 1996). nitive processes. Based on the findings of previous studies To be able to measure what happens when Type 1 pro- (Benton & Burgess, 2009; Edmonds & Jeffes, 2009, Booth cessing is in conflict with the more reflective Type 2 pro- et al., 2012, Edmonds et al. 2013, Rogers et al., 2001, Benton cessing, Frederick (2005) developed the CRT. This brief et al., 2016), we expected performance for sustained atten- test consists of short maths puzzles, which can appear at tion and executive function tests to be positively affected first glance as having a very obvious answer. However, the by water consumption (Hypothesis 1). It was predicted that respondent’s initial intuitive response is incorrect, and they since water consumption has positive effects on typical cog- can only arrive at the correct response if they suppress their nitive processing tests, this should extend to judgment and initial heuristic answer and engage in more reflective think - decision-making tasks (Hypothesis 2) and that attention and ing. The CRT has been linked to performance on numerous executive performance scores (here: set shifting) correlate judgment and decision-making tasks, such as risk taking, with judgement and decision-making tasks (Hypothesis 3). temporal discounting, and use of heuristic thinking (Fred- If so, judgement and decision-making scores may be dif- erick, 2005; Toplak, Stanovich, & West, 2011). ferentially associated—depending on water supplementa- Hydration has been previously linked to activation in tion—with scores of tasks measuring general levels of atten- the anterior cingulate cortex (ACC), and the latter in turn tion (Hypothesis 4a), mental manipulation of information has been proposed to be an important component of frontal (Hypothesis 4b) or inhibition performance (Hypothesis 4c). attentional control systems (Braver, Barch, Gray, Molfese, The last two predictions are mainly based on the dual pro- & Snyder, 2001; Bush et al., 1998; Pardo, Pardo, Janer, & cess theories predicting monitoring of type 1 processing by a 1 3 Psychological Research reflective system that either inhibits automatic thinking pro- et al., 2013). This was converted to a percentage where a cesses (Frederick, 2005; Kahneman, 2011), increases think- higher percentage indicated a higher level of thirst. ing performance in terms of mental manipulation abilities, or both (Toplak et al., 2011). Cambridge neuropsychological test automated battery Importantly, we controlled for whether these effects were (CANTAB) moderated by thirst or physiological dehydration status. Uri- nary osmolality analysis was used to assess hydration status The CANTAB eclipse software (Sahakian & Owen, 1992) to formally examine a potential association between hydra- contains an array of tests used to assess cognitive perfor- tion status, thirst, drinking water and cognitive performance. mance. We administered six tests from this platform: the motor screening test (MOT), the simple reaction time (SRT), the choice reaction time (ChoiceRT), the big/little circle (BLC), the intra-extra dimensional set shifting (IED), and Methods the rapid visual information processing (RVP). The IED, RVP and ChoiceRT assess executive functions including Participants visual attention, which is the focus of this report. ChoiceRT is a 2-choice reaction time test with stimulus 31 participants (14 males) were recruited for this study and response uncertainty introduced by having two possi- through advertisements placed at the University of East ble stimuli and two possible responses. Participants were London (UEL) and on psychology websites, via emails instructed to press the left-hand button if the stimulus (an to UEL students, and through friends. A pre-participation arrow) was displayed on the left-hand side of the screen, health questionnaire was sent to interested individuals, to and the right-hand button if the stimulus was displayed on exclude persons for whom overnight fasting might have been the right-hand side of the screen. A practice stage (24 trials) a potential health risk (e.g., pregnant women and people was followed by two assessment stages (50 trials each). The suffering from diabetes or a heart condition). The ages of dependent variable was reaction time. participants ranged from 21 to 46 years (mean 31 years; IED is a test of rule acquisition and reversal. Two pat- SD 7.24). terns are displayed on the screen, first simple (colour-filled shapes) and then compound (white lines overlying colour- Tasks and questionnaires filled shapes). The participant must learn which of the two stimuli was correct by trial and error learning. When six Task were presented to participants in the following order. consecutive correct responses were recorded, the contingen- Parallel forms of all tasks were administered counterbal- cies were reversed and this pattern of stimulus addition and anced across water condition and task version. reversal continued for nine blocks. If the participant failed to reach six consecutive responses after 50 trials, the test was terminated. The dependent variable for this task was the total International positive and negative affect schedule errors committed. short form (I-PANAS-SF) RVP is a sensitive measure of general performance and in particular of visual sustained attention. Numbers appear The I-PANAS-SF scale is a shorter version of the original one at a time in a box in the centre of the screen at the rate of PANAS consisting of 10 items instead of 20 (Thompson, 100 digits per minute. Participants were instructed to press 2007) used to measure general affect. Half of the emotion the button on the press pad whenever they spotted a target words presented reflect negative affect states (ashamed, sequence of three consecutive numbers. A practice stage afraid, hostile, nervous, upset) and the other half reflect (lasting 2 min) in which participants were prompted as to positive affect states (active, alert, attentive, determined, when a sequence had begun and when to press the button inspired). Participants rated their positive and negative affect was followed by a test stage (lasting 4 min) in which no cues on a 5-point scale that ranged from “very slightly or not at were displayed and the participant had to spot three different all” (1) to “extremely” (5). sequences on their own. Target sequences occurred at the rate of 16 every 2 min. The measured dependent variable Thirst scale was total error rate. Participants were asked to indicate their level of thirst by Measuring cognitive reflection performance marking an X on a continuous horizontal line (17.8 cm) with anchors indicating “not at all” to “very thirsty” (Edmonds To assess judgement and decision-making performance, we employed tasks that are typically used to assess the use 1 3 Psychological Research of heuristic (automatic) processing that can be overcome long would it take for the patch to cover half of the lake? by reflective (controlled and analytic) thinking (follow - (Intuitive answer: 24; correct answer: 47). ing largely Toplak et al., 2011). This consisted of nine vignettes or puzzles in total per session. Six of these were Measurement of hydration status heuristics-and-biases vignettes from widely cited publica- tions that reflect important aspects of rational thought such A Vitech Advanced Multi Sample Micro freezing point as probabilistic reasoning, hypothetical thought, theory osmometre from Advanced Instruments Inc. was used to justification, scientific reasoning, and the tendency to think determine urine osmolality (mOsm/kg) to assess partici- statistically. Each answer to a heuristic vignette task was pants’ hydration status. A higher value indicates a greater scored as correct or incorrect (1 or 0 score), resulting in a degree of dehydration. According to the US National Insti- total maximum score of 6 (per session). The battery was tutes of Health, a concentration of 500–800 mOsm/kg is comprised of the following: considered normal, whereas a 12–14 h fluid restriction should yield a value in excess of 850 mOsm/kg (Chern- 1. Causal base rate (Fong, Krantz, & Nisbett, 1986). ecky & Berger, 2012). A higher value indicates a greater 2. Sample size (Tversky & Kahneman, 1974). degree of dehydration. 3. Gambler’s fallacy (Toplak et al., 2011). 4. Conjunction fallacy (Tversky & Kahneman, 1983). 5. Bayesian reasoning (Doherty & Mynatt, 1990). Procedure 6. Sunk cost (Arkes & Blumer, 1985). A pre-participation health questionnaire was sent to inter- Example of sample size: ested individuals, to exclude persons for whom overnight A certain town is served by two hospitals. In the larger fasting might have been a potential health risk (e.g., preg- hospital about 45 babies are born each day, and in the nant women and people suffering from diabetes or a heart smaller hospital about 15 babies are born each day. As you condition). They were also provided with an empty sample know, about 50% of all babies are boys. However, the exact container in which they supplied their waking urine sam- percentage varies from day to day. Sometimes it may be ple, which they also brought with them to each of their higher than 50%, sometimes lower. For a period of 1 year, sessions. each hospital recorded the days on which more than 60% Participants visited UEL’s Psychology Research Suite of the babies born were boys. Which hospital do you think on two occasions, 1 week apart, after having fasted (no recorded more such days? food or drink) from 9 p.m. the night before. Participants were asked to collect a urine sample upon waking (in ster- (a) The larger hospital. ile sample pots already provided), which they brought with (b) The smaller hospital. them. Testing took place in the mornings (8 a.m.–11 a.m.). (c) About the same (that is, within 5% of each other). To standardise the water content of breakfast, before each testing session, participants received a choice of cereal bar In addition to the vignettes inducing heuristic thinking, (113 kcal or 119 kcal). On one occasion (counterbalanced the Cognitive Reflection Test (CRT; Frederick, 2005) was across participants), they were also given a 500 ml bottle used. The CRT is designed to measure participants’ ten- of water (at room temperature). Participants were explic- dency to override an intuitive first response and to engage itly and clearly instructed to drink as much as they wanted in reflective thinking to arrive at the correct answer (simi- before beginning the tasks. There was no time pressure, lar to the mechanism proposed to work in solving heuris- but all participants stopped drinking after 2  min. They tic vignettes, Kahneman, 2011). The dependent variable were not allowed to continue drinking during testing. was the total number of correct responses (maximum of Participants then completed the tasks in the order they 3 per session). The original CRT comprised of only three have been described above. At the end of testing, they questions. We used the extended version by Toplak et al. were asked to provide another urine sample. The second (2014) resulting in different three questions in each of the session followed the same procedure and at the end of the two sessions. The answers to the six heuristic vignettes second session they were debriefed and compensated for and the three CRT puzzles formed the cognitive reflection their time and participation. Tasks in both sessions were score (a maximum of nine correct answers per session). completed in approximately 1 h. An example of the CRT is the following: in a lake, there is The order of water supplementation and tasks admin- a patch of lily pads. Every day, the patch doubles in size. If istered was counterbalanced so that 15 participants had it takes 48 days for the patch to cover the entire lake, how water in their first session and 14 in their second session, and 15 had version A of decision-making tasks in their 1 3 Psychological Research first session and 14 had version B of decision-making tasks Water consumption and hydration status effects in their second session. on thirst and mood scales Data analysis In the water condition, participants drank a mean of 303.44 ml (SD 158.21; range 50–500 ml). To test whether The main aim of this study was to investigate the effect of people were indeed dehydrated in the no water condition water supplementation on cognitive performance. To test after test as well as on both mornings, we ran a 2 (water vs hypothesis 1 and 2, the data was subjected to a series of no water) by 2 (waking vs end of test) ANOVA on osmolal- mixed analyses of variance (ANOVA) in which water sup- ity readings. There was no effect of day, F < 1, but a (1,28) plementation (water/no water given) was a within-partic- main effect of test time, (F = 5.96, p = .021, η = 0.176), (1,28) p ipants factor, and order (water first/no water first), thirst as well as an interaction, (F = 6.231, p = 0.019, (1,28) (thirsty/not thirsty), and urine osmolality (high/low) were η = 0.182). Whereas there was no difference between between-participants factors. The same analyses were also hydration readings in the water condition before (M = 735, performed for the combined cognitive ref lection scores. SD = 252) and after (M = 758, SD = 235) testing, there was For thirst and hydration, median splits were performed a difference in the no water day, with readings lower before grouping participants as either thirsty/not thirsty and (M = 693, SD = 218) than after (M = 813, SD = 217) testing hydrated/not hydrated based on the respective medians of (see Fig.  1). This suggests that on a group level, partici- 63% and 827.5 mOsm/kg on the ‘no water day’. The post- pants were reasonably dehydrated (osmolality readings of ca test osmolality data was used in the present analyses. The 700–800 mOsmo/kg), but also that in the no water day, the pre-test data was used to confirm fasting (see “Results”). dehydration became significantly worse during the morning To investigate hypothesis 3 and 4, correlation analy- compared to the water day (Edmonds et al., 2013). Thus, ses were also performed in an attempt to tease apart a water supplementation on the water day prevented further possible relationship between performance on the judg- dehydration, which seemed to happen on the no water day ment and decision-making tasks and performance on the as testing went on through the morning. Thirst ratings also ChoiceRT, IED, and RVP. confirmed that participants arrived thirsty: participants rated themselves as having greater subjective thirst on the occasion that they were not offered water (F = 46.112, (1,27) p < 0.001). Results The responses to the I-PANAS-SF mood scale were mostly unaffected by water supplementation, thirst, order Data from two participants could not be analysed because and osmolality. There were two exceptions to this state- they did not return for the second session. The final sam- ment: there was a water supplementation x order interac- ple size was 29 participants (16 females). tion for “attentive” and an osmolality effect on “inspired”. Fig. 1 Mean osmolality readings for participants on different sessions and time points during testing days (before—“waking”—and after tests). Error bars are standard errors of the mean. Significant differences (using “asterisk” to denote p > .05) between condi- tions are indicated 1 3 Psychological Research Participants who received water in their first session factor), again found similar effects for the independent vari - reported being more “attentive” on that occasion compared ables on CANTAB scores, all Fs < 1 (except the trend of to their second session in which they did not have any water water for ChoiceRT, with p values between 0.065 and 0.071. (F = 16.00, p < 0.001). In the case of urine osmolal- (1,27) ity, dehydrated participants (as evidenced by higher urine Water consumption effects on judgment osmolality) rated themselves as significantly less “inspired” and decision‑making performance (F = 4.276, p = 0.048). There was no effect of thirst (1,27) (high vs low scorers) on any of the items presented in the Three mixed-design ANOVAs were performed with total I-PANAS-SF mood scale. correct score for the combined judgement and decision-mak- ing tasks (six heuristic vignettes and three CRT vignettes in Water consumption effects on executive functions each session, see Methods) as the dependent variable, analo- gous to the ANOVAs for the executive function tests above. Mean scores on CANTAB tests were screened for normal The within-participant factor in each ANOVA was water distribution and outliers, using the interquartile range rule of supplementation (water vs no water). The between factors g = 3 (Hoaglin et al, 1986). Only one RVP errors data point in the respective ANOVAs were order (water first session or was substituted with RVP misses in one condition for one water second session), hydration (osmolality: high or low; participant who had a very high RVP false alarm rate in one i.e., dehydrated or hydrated), and finally thirst (high or low condition. For all other participants, the RVP total errors after median split). In all three ANOVAs, there was a main were calculated as the sum of the number of false alarms effect of water supplementation (Table  4), for the ANOVA and number of misses. on water and order (F = 7.37, p = 0.011, η = 0.215), (1,27) p Performance on each of the CANTAB tasks was analysed water and hydration (F = 7.44, p = 0.013, η = 0.209), (1,27) p using mixed-design ANOVAs, one separately for effects of and water and thirst (F = 7.69 p = 0.012, η = 0.212). (1,27) p order, thirst, and osmolality. The within-participants factor Participants scored overall higher on the judgment and deci- in each ANOVA was water supplementation (water vs no sion-making tasks in conditions in which they received water water). The between factors in the respective ANOVAs were compared to the no water day (Fig. 2). There were no simple order (water r fi st session or water second session), hydration main effects from factors order, F 2.730, p = .110, hydra- (1,27) [osmolality: high > 827.5 mOsm/kg or low < 827.5 mOsm/ tion F < 1, or thirst, F = 2.33, p = .138. There were (1,27) (1,27) kg; i.e., dehydrated (15) or hydrated (14)], and finally thirst also no interaction effects involving order, all F s < 1. Water (high or low after median split; 14 participants classified supplementation therefore had a positive effect on scores as thirsty and 15 as not thirsty). There were no significant across the battery of judgment and decision-making tasks, effects or trends for the factor order or water (see Tables  1, 2, relatively independent of levels of thirst and hydration (on 3), bar two exceptions. There were trends for ChoiceRTs to the no water day), or order. The ANCOVAs using thirst and be generally faster in the water conditions (p values between osmolality (mean-centred) as co-variates instead of median 0.066 and 0.073; Tables 1, 2, 3), and there was a significant splitting found the same patterns effects on cognitive reflec- interaction for water and order in the RVP tasks (Table 1), tion scores, with no main or interaction effect of the co- with more errors in the water condition (M = 17.80, variates (all Fs < 1). SD = 4.72) compared to the no water condition (M = 21.13, This result confirmed hypothesis 2, that water supplemen- SD = 5.55) when participants received water in their first ses- tation increased cognitive reflection scores, and this result sion, p = 0.007, but vice versa when they received it second, was not qualified by any interaction. p = 0.057 (water: M = 22.07, SD = 3.15; no water: M = 20.50, SD = 3.65). Therefore, hypothesis 1 could not be retained. Correlation analysis Regarding the main effects of between-subjects vari- ables, there was a marginal effect of order on the ChoiceRT, To investigate the possible relationship between hydration F = 4.034, p = .055, with higher RTs in the first session variables, judgment scores and executive functions for dif- (1,27) (M = 327, SD = 12) than in the second (M = 290, SD = 13). ferent water supplementation conditions, we performed cor- Otherwise, there were no effects, all Fs < 1, except for IED relation analyses. Table 5 shows differing degrees of associa- errors and order, F = 1.503, p = .23, and order effects on tions depending on whether the data used was taken from the (1,27) RVP error rates, F = 2.161, p = .153. Controlling for the day participants received water or not. (1,27) amount of water each participant drank (using ANCOVAs) There were significant correlations between cognitive did not change the pattern of effects, all F s < 1, except reflection scores and ChoiceRT (water: r = − 0.473, p = (1,28) for a similar trend as above for ChoiceRT F = 3.53, 0.010; no water: r = − 0.579, p = 0.001) and IED errors (1,28) p = .71. Additional ANCOVAs using thirst and osmolality (water: r = − 0.533, p = 0.003; no water: r = − 0.578, as co-variates (rather than median split as a between-group p = 0.001) on both days, water and no water, respectively. 1 3 Psychological Research Fig. 2 Mean scores for the com- bined judgment and decision- making tasks comparing perfor- mance on the day participants received water with the day they did not. Water consumption has a significant effect on scores, with participants scoring better on the day they did receive water Table 1 CANTAB test means, SDs and F ratios by water condition (water/no water) and order (water first/no water first) Task Water first No water first Results from the omnibus statistical analysis; those with p < .05 in bold Water No water Water No water M SD M SD M SD M SD ChoiceRT 317.81 51.63 336.33 74.21 288.73 31.55 292.27 35.05 Water F = 3.476, p = 0.073 (1,27) Water × order F = 1.600, p = 0.217 (1,27) IED total errors 20.07 10.53 18.33 11.76 13.79 9.31 15.43 9.99 Water F = 0.002, p = 0.966 (1,27) Water × order F = 2.615, p = 0.117 (1,27) RVP errors 10.53 5.64 7.60 6.34 5.50 3.43 7.42 4.21 Water F = 0.335, p = .568 (1,26) Water × order F = 14.259, p = 0.001 (1,26) Table 2 CANTAB test means, SDs and F ratios by water condition (water/no water) and post-testing urine osmolality (low/high) as measured on the day participants did not receive any water Task Low osmolality High osmolality Results from the omnibus statistical analysis; those with p < .05 in bold Water No water Water No water M SD M SD M SD M SD ChoiceRT 304.26 46.83 319.05 75.54 303.31 44.61 311.34 48.19 Water F = 3.548, p = 0.070 (1,27) Water × osmo F = 0.312, p = 0.581 (1,24) IED total errors 17.53 11.04 18.67 11.32 16.50 9.80 15.07 10.41 Water F = 0.019, p = 0.891 (1,27) Water × osmo F = 1.446, p = 0.240 (1,24) RVP errors 8.57 5.71 7.64 5.51 7.71 5.21 7.50 5.53 Water F = 0.552, p = 0.464 (1,26) Water x Osmo F = 0.169, p = 0.684 (1,26) There was also a significant correlation between CRT executive function tasks being associated with higher cogni- scores and RVP errors on the water day only (r = − 0.451, tive reflection performance. Hypothesis 3 was therefore con- p = 0.014). All correlations were in the predicted direction firmed—performance on executive function tasks (though with better performance (lower errors or shorter RTs) in only in the water condition for RVP) was associated with 1 3 Psychological Research Table 3 CANTAB test means, SDs and F ratios by water condition (water/no water) and thirst (low/high) as measured on the day participants did not receive any water Task Low thirst High thirst Results from the omnibus statistical analysis; those with p < .05 in bold Water No water Water No water M SD M SD M SD M SD ChoiceRT 303.96 41.45 309.75 38.81 303.57 49.87 320.75 80.94 Water F = 3.676, p = 0.066 (1,27) Water × thirst F = 0.903, p = 0.350 (1,24) IED total errors 16.29 10.61 15.43 10.17 17.73 10.31 18.33 11.62 Water F = 0.011, p = 0.916 (1,27) Water × thirst F = 0.124, p = 0.728 (1,27) RVP errors 8.14 5.75 7.64 4.73 8.14 5.20 7.50 6.21 Water F = 0.536, p = 0.470 (1,27) Water × thirst F = 0.024, p = 0.879 (1,27) Table 4 Cognitive reflection Water No water Results from the statistical analysis score means, SDs and F ratios by water condition (water/ M SD M SD no water) and post-testing Water first 4.80 2.18 4.27 1.62 Water F = 7.374, p = 0.011, η = 0.215 urine osmolality (low/high) (1,27) p as measured on the day Water second 6.00 1.47 5.07 1.77 Water × order F = 0.539, p = 0.469 (1,27) participants did not receive any 2 Osmo low 5.47 2.44 4.53 1.92 Water F = 7.439, p = 0.013, η = 0.209 (1,27) p water Osmo high 5.29 1.27 4.79 1.53 Water × osmo F = 0.651, p = 0.427 (1,27) Thirst low 4.87 2.13 4.27 1.53 Water F = 7.688, p = 0.012, η = 0.212 (1,27) p Thirst high 5.93 1.59 5.07 1.86 Water × thirst, F = 0.226, p = 0.639 (1,27) Table 5 Correlations between cognitive reflection performance vignettes. Some approaches in the dual systems framework scores and water consumption, urine osmolality, thirst, CANTAB (e.g., Evans & Stanovich, 2013) further implicate mental tasks and for both days (participants received/did not receive water) simulation performance, the ability to maintain and sym- (N = 29) bolically manipulate separate mental representations of a Measured variable Water No water problem. ChoiceRT latency difference scores were log-transformed Water consumed (ml) 0.208 – to reduce potential issues of positive skew and normality of Urine osmolality (post) 0.132 0.107 residuals. Results of the multiple linear regression indicated Thirst −0.168 0.169 that there was a combined significant effect of differences in ChoiceRT (RT) −0.473** −0.579*** ChoiceRT and IED (errors) explaining differences in cogni- IED (errors) −0.533** −0.578** tive reflection scores, (F = 3.765, p = 0.037, R = 0.224). (2,26) RVP (errors) −0.451* −0.148 ChoiceRT difference (t = − 2.244, p = 0.034) was a signifi- *p ≤ 0.05. **p < .01. ***p < .001 cant predictor in the model, but not IED error difference (t = − 1.543, p = 0.135) (Table 6). Adding RVP errors (difference higher performance in the judgment and decision-making scores) as a predictor variable again showed a relationship tasks. for cognitive reflection scores with ChoiceRT (t = − 2.343, Finally, a linear regression analysis was performed p = 0.027) but not RVP errors (t = − 1.005, p = 0.324), with using difference scores (water–no water). The difference the overall model marginally significant, (F = 4.899, p = (2,25) in cognitive reflection scores between the no water and the 0.058, R = 0.254). Thus, hypothesis 4c was retained: cogni- water condition served as the dependent variable (crite- tive reflection scores for sessions in which water was given rion) and the difference (between water and no water day) were differentially influenced by ChoiceRT scores compared in ChoiceRT and IED errors as independent variables. The to sessions in which water was not given—the higher the regression model tested whether differences (between ses- differences in ChoiceRT latencies (and therefore the worse sions) in the executive function tasks were associated with the inhibition performance between no water and water con- differences in the cognitive reflection scores. Recall that dition), the lower the improvement of cognitive reflection the hypothesis was based on the premise by dual process scores from no water to water condition. theories that increased inhibition processes are related to Hypothesis 4a and 4b were therefore not retained—dif- increased performance in CRT-like puzzles and heuristic ferences in tasks measuring attention performance (RVP) or 1 3 Psychological Research Table 6 Summary of regression analysis for variables predicting cognitive reflection score differences (N = 29) between sessions Source B SE B β t p LBCI 95% UBCI 95% ChoiceRT— −4.357 1.942 −0.388 −2.244 0.034* −8.349 −0.366 diff IED (errors)— −0.066 0.043 −0.267 −1.543 0.135 −0.155 0.022 diff mental simulation (IED) between the water supplementation (Edmonds et al., 2013). However, we have failed to replicate conditions were not associated with the difference in cogni - the significant findings pertaining to participants’ subjective tive reflection scores. ratings of thirst as a moderator of the effects of water supple- mentation on most the measures assessed (including mood ratings). This could be an idiosyncrasy of the particular sam- Discussion ple population or it could indicate individual differences in feelings of subjective thirst. For example, several partici- The current study is to our knowledge the first to report pants (N = 7) in this study spontaneously expressed that they increased cognitive reflection performance (and, by exten - were seldom thirsty, so it was perhaps not surprising that sion, increased judgment and decision-making performance, even on the occasion when they were not given any water, Frederick, 2005; Toplak et al., 2011) after water consump- they indicated a relatively low level of thirst. Nonetheless, tion. When thirst, hydration status, and mood state were con- as elucidated by the relevant statistical analysis on a group trolled for, water supplementation increased performance on level, participants did report experiencing greater levels of an overall composite score from widely used judgment and subjective thirst on the occasion they did not receive any decision tasks (judgment vignettes eliciting heuristic think- water. ing, simple maths puzzles requiring cognitive reflection). Our main result is the significant effect of water supple- These scores were related to inhibition processing speed and mentation on performance on the judgement and decision- executive functions (ChoiceRT and IED), but not attentional making tasks (heuristics and biases, cognitive reflection performance (RVP) or feelings of general attentiveness. The test)—participants performed better on the occasion on experimentally induced differences in judgement and deci- which they received water. This finding cannot be easily sion performances between water days and no water days dismissed as a result of demand characteristics (i.e., simply were associated with differences in Stroop-like task perfor - being given a drink increasing motivation, or expectation of mances (and Simon task) generally associated with inhibi- doing better), because we did not find an influence of water tion processes. Before we turn to these effects in detail, we supplementation on mood effects (see also Edmonds et al., discuss the physiological factors that could have influenced 2013 who show that expecting water supplementation does this result. not explain increased performance in attention tasks) nor on In general, there were no effects of water, thirst or hydra- other cognitive tasks (IED, RVP). Therefore, we interpret the tion status (except for PANAS ‘attentive’ scores, but those effects on cognitive reflection scores as substantially driven were not associated with cognitive reflection performance) by water supplementation. on the measures of mood used in this study. Some stud- The tasks used here were aimed at assessing ‘slow’ pro- ies have previously reported links between dehydration cessing (reflective thinking) vs ‘fast’ processing (heuristic and mood ratings (Shirreffs et al., 2004), and water supple- thinking; Kahneman, 2003), with the particular aim to inves- mentation and mood ratings (Edmonds et al., 2013), while tigate potential processes that override decisions reached by others report that water supplementation does not affect Type 1 (De Neys & Glumicic, 2008). Inhibition performance mood (Edmonds et al., 2013). Furthermore, it may be that has been shown to be influenced by water supplementation whether mood affects dehydration may depend on the man- in previous research (Edmonds et al., 2013) and could thus ner in which dehydration is achieved: Shirreffs et al. (2004) modulate the effect of water supplementation on cognitive induced dehydration by fluid restriction, whereas Edmonds reflection performance. Indeed, performance on the Stroop- et  al. (2013) reported effects on water supplementation. like ChoiceRT task correlated with judgment and decision At any rate, the main finding here is that mood (and hence performance score in both conditions, as was predicted expectation effects) does not explain the findings for the (and replicated previous results, e.g., Toplak et al., 2011). effect of water on cognitive reflection tasks. In addition, regression analysis suggests a link between inhi- Previous studies have revealed that both water sup- bition performance (as measured by the ChoiceRT) and the plementation and thirst impact on cognitive performance effect of water supplementation on decision performance: 1 3 Psychological Research As ChoiceRT performance is affected by supplementation 2000). Similarly, Evans and Stanovich (2013) propose that (although effect sizes are small), so is the performance on the reflective system requires executive processes beyond the heuristic vignettes and puzzles. Of course, we cannot inhibition to enable ‘cognitive decoupling’, that is the directly infer causation, but it is noteworthy that most dual ability for mental simulation and abstract thinking. It is process theories of thinking and deciding (De Neys & Glu- this ability that potentially allows the independent mental micic, 2008; Evans & Stanovich, 2013; Kahneman, 2011) representation of information in math-like puzzles (CRT) predict that cognitive reflection performance relies on moni- and vignettes (heuristics) shown to participants. Indeed, toring and consequently inhibiting the pre-potent responses our findings of strong correlations between IED (set shift- related to heuristic thinking. It is the successful monitoring ing) and cognitive reflection tasks strongly indicates that and inhibition that consequently decreases biased judgments. some form of cognitive decoupling underlies Type 2-like Our finding that particular executive processes correlate processing. But here again, there was no indication that with cognitive reflection tasks is also roughly in line with IED—as a proxy measure for mental simulation- modu- the psychobiological literature, especially the notion of a lates the relationship between water supplementation and possible role of a behavioural inhibition in judgement per- cognitive reflection scores, unlike what we found with the formance. For example, fMRI studies found that dehydra- ChoiceRT task. tion directly affects the blood flow to the anterior cingulate Our findings are therefore the first that show a tentative cortex (Farrell et al., 2008), which is linked to inhibition link between water supplementation (after dehydration), (e.g., Stroop) performance. When comparing incongruent inhibition performance, and judgement and decision- and control conditions, the majority of such studies report making processes. If future research confirms the effect maximal differential activation occurring in the anterior cin- of water supplementation (and a possible role of dehy- gulate cortex (Bench et al., 1993; Bush at al., 1998; Carter dration) on decision-making performance, the underlying et al., 1995; Carter et al., 2000; Derbyshire, Vogt, & Jones, cognitive–physiological mechanisms may be more com- 1998; Pardo et al., 1990). Although this area shows great- plicated. Hydration has been linked to a range of inhibi- est activation in the incongruent condition of the Stroop, tory or excitatory effects, leading to cognitive improve - the congruent condition (facilitation) has also been shown ments or impairments. For example, chronic dehydration to increase activation as compared to a control condition in animals increases the release of the neurotransmitters (Bench at al., 1993; Carter et al., 1995). Thus, even though gamma-aminobutyric acid (GABA) and glutamate, which the choice response times may be deemed a somewhat indi- have inhibitory and excitatory effects, respectively (Di & rect inhibition measure, both congruent and incongruent Tasker, 2004). Furthermore, dehydration has been shown trials (and latency data) may indicate inhibition processes. to increase the release of the stress hormone cortisol Furthermore, previous work has shown that complex pro- (Francesconi et al., 1984), and elevated cortisol levels have cessing speed measures are substantially correlated with been associated with impaired cognitive function (Green- executive control measures but not with simpler speed meas- dale et al., 2000; Kirschbaum et al., 1996). In the present ures (e.g., Cepeda et al., 2013). However, future research data, similar conflicting effects may therefore account needs to further elucidate the exact mechanisms of inhibition for the difficulty in establishing stronger links between and facilitation linked to water supplementation. Moreover, executive functions and cognitive reflection performance different tasks may tap into different inhibition processes. in different water supplementation (and hence hydration) For example, Khng and Lee (2014) found performance on conditions. the Stroop tasks largely independent from performance on a In conclusion, we find a clear effect of water supple- Stop-signal task, indicating potentially different underlying mentation (after dehydration) on decision-making per- inhibition processes. Although the ChoiceRT task employed formance when thirst is controlled for. The challenge here contains conditions that require Stroop-like inhibition for future studies will be to further clarify the relation- processes, other tasks may help establish better models ship between physiological and cognitive mechanisms. explaining the relationship between inhibition, executive Researchers will need to employ executive function tests functions, and cognitive reflection performance. In any that are sensitive to different types of inhibition mecha- case, though intriguing, the link between a task involving nisms and other executive processes, as well as measuring inhibition processes and cognitive reflection performance effects stemming from physiological hydration and thirst. found here needs to be interpreted with caution until further Acknowledgements We thank Dr. Caroline Edmonds for her com- replicated with other measures of inhibition. ments on the manuscript. We would like to thank Dr. Paula Booth for In addition to inhibition, Toplak et al. (2011) found that her help and support in running this study, and for her guidance with the cognitive reflection task (CRT) also correlates with regard to the urine osmolality analysis in particular. measures of cognitive ability (see also Stanovich & West, 1 3 Psychological Research Funding This work was supported by a student grant to Olivia Patsalos https ://doi.or g/10.1002/(SICI)1097-0193(1998)6:4%3C270 from the European Hydration Institute. ::AID-HBM6%3E3.0.CO;2-0 Carter, C. S., Macdonald, A. M., Botvinick, M., Ross, L. L., Stenger, V. A., Noll, D., & Cohen, J. D. (2000). Parsing executive processes: Compliance with ethical standards Strategic vs. evaluative functions of the anterior cingulate cortex. Proceedings of the National Academy of Sciences of the United Conflict of interest Dr. Volker Thoma declares that he has no conflict States of America, 97(4), 1944–1948. https ://doi.org/10.1073/ of interest. Olivia Patsalos declares that she has no conflict of interest. PNAS.97.4.1944. Carter, C. S., Mintun, M., & Cohen, J. D. (1995). Interference and Ethical approval All procedures performed in studies involving human facilitation effects during selective attention: An H215O PET participants were in accordance with the ethical standards of the insti- study of Stroop task performance. NeuroImage, 2(4), 264–272. tutional and/or national research committee and with the 1964 Helsinki https ://doi.org/10.1006/nimg.1995.1034. declaration and its later amendments or comparable ethical standards. Cepeda, N. J., Blackwell, K. A., & Munakata, Y. (2013). Speed isn’t everything: Complex processing speed measures mask individ- Informed consent Informed consent was obtained from all individual ual differences and developmental changes in executive control. participants included in the study. Developmental Science, 16(2), 269–286. Chernecky, C., & Berger, B. (2012). Laboratory tests and diagnostic procedures. 6th edn. Saunders. eBook ISBN: 9781455745029. Open Access This article is distributed under the terms of the Crea- Cokely, E. T., & Kelley, C. M. (2009). Cognitive abilities and supe- tive Commons Attribution 4.0 International License (http://creat iveco rior decision making under risk: A protocol analysis and process mmons.or g/licenses/b y/4.0/), which permits unrestricted use, distribu- model evaluation. Judgment and Decision Making, 4(1), 20–33. tion, and reproduction in any medium, provided you give appropriate https ://doi.org/10.1016/j.jbank fin.2009.04.001. credit to the original author(s) and the source, provide a link to the De Neys, W., & Glumicic, T. (2008). Conflict monitoring in dual pro- Creative Commons license, and indicate if changes were made. cess theories of thinking. Cognition, 106(3), 1248–1299. https :// doi.org/10.1016/j.cogni tion.2007.06.002. Derbyshire, S. W. G., Vogt, B. A., & Jones, A. K. P. (1998). Pain and Stroop interference tasks activate separate processing modules in References anterior cingulate cortex. Experimental Brain Research, 118(1), 52–60. https ://doi.org/10.1007/s0022 10050 254. Arkes, H. R., & Blumer, C. (1985). The psychology of sunk cost. Di, S., & Tasker, J. G. (2004). Dehydration-induced synaptic plasticity Organizational Behavior and Human Decision Processes, 35(1), in magnocellular neurons of the hypothalamic supraoptic nucleus. 124–140. https ://doi.org/10.1016/0749-5978(85)90049 -4. Endocrinology, 145(11), 5141–5149. h t tp s : // d o i. o r g /1 0 .1 2 10 / Bar-David, Y., Urkin, J., & Kozminsky, E. (2005). The effect of vol- en.2004-0702. untary dehydration on cognitive functions of elementary school Document_not_found. (2018). Document not found (Frederick, 2005). children. Acta Paediatrica, 94(11), 1667–1673. h t t p s : / / d o i . Doherty, M. E., & Mynatt, C. R. (1990). Inattention to P(H) and to org/10.1080/08035 25050 02546 70. P(D/~H): A converging operation. Acta Psychologica, 75(1), Bench, C. J., Frith, C. D., Grasby, P. M., Friston, K. J., Paulesu, E., 1–11. https ://doi.org/10.1016/0001-6918(90)90063 -L. Frackowiak, R. S., & Dolan, R. J. (1993). Investigations of the Edmonds, C. J., & Burford, D. (2009). Should children drink more functional anatomy of attention using the Stroop test. Neuropsy- water? The effects of drinking water on cognition in children. chologia, 31(9), 907–922. Appetite, 52(3), 776–779. https ://doi.or g/10.1016/j.appe t Benton, D. (2011). Dehydration influences mood and cognition: .2009.02.010. A plausible hypothesis? Nutrients, 3, 555–573. https ://doi. Edmonds, C. J., Crombie, R., Ballieux, H., Gardner, M. R., & Dawkins, org/10.3390/nu305 0555. L. (2013). Water consumption, not expectancies about water con- Benton, D., & Burgess, N. (2009). The effect of the consumption of sumption, ae ff cts cognitive performance in adults. Appetite, 60(1), water on the memory and attention of children. Appetite, 53(1), 148–153. https ://doi.org/10.1016/j.appet .2012.10.016. 143–146. https ://doi.org/10.1016/j.appet .2009.05.006. Edmonds, C. J., & Jeffes, B. (2009). Does having a drink help you Benton, D., Jenkins, K. T., Watkins, H. T., & Young, H. A. (2016). think? 6–7-year-old children show improvements in cognitive per- Minor degree of hypohydration adversely influences cognition: A formance from baseline to test after having a drink of water. Appe- mediator analysis. American Journal of Clinical Nutrition, 104(3), tite, 53(3), 469–472. https://doi.or g/10.1016/j.appet.2009.10.002 . 603–612. https ://doi.org/10.3945/ajcn.116.13260 5. Evans, J. S. B. T. (2003). In two minds: Dual-process accounts of rea- Booth, P., Taylor, B., & Edmonds, C. (2012). Water supplementation soning. Trends in Cognitive Sciences, 7(10), 454–459. https://doi. improves visual attention and fine motor skills in school children. org/10.1016/j.tics.2003.08.012. Education and Health, 30(3), 75–79. Evans, J. S. B. T. (2008). Dual-processing accounts of reasoning, judg- Braver, T. S., Barch, D. M., Gray, J. R., Molfese, D. L., & Snyder, A. ment, and social cognition. Annual Review of Psychology, 59(1), (2001). Anterior cingulate cortex and response conflict: Effects of 255–278. https://doi.or g/10.1146/annure v.psych.59.10300 6.09362 frequency, inhibition and errors. Cerebral Cortex, 11(9), 825–836. https ://doi.org/10.1093/cerco r/11.9.825. Evans, J. S. B. T., & Stanovich, K. E. (2013). Dual-process theories Buckner, R. L., & Carroll, D. C. (2007). Self-projection and the of higher cognition: Advancing the debate. Perspectives on Psy- brain. Trends in Cognitive Sciences, 11(2), 49–57. https ://doi. chological Science, 8(3), 223–241. https://doi.or g/10.1177/17456 org/10.1016/j.tics.2006.11.004. 91612 46068 5. Bush, G., Whalen, P. J., Rosen, B. R., Jenike, M. A., McInerney, S. Fadda, R., Rapinett, G., Grathwohl, D., Parisi, M., Fanari, R., Calò, C. C., & Rauch, S. L. (1998). The counting Stroop: An interference M., & Schmitt, J. (2012). Effects of drinking supplementary water task specialized for functional neuroimaging–validation study at school on cognitive performance in children. Appetite, 59(3), with functional MRI. Human Brain Mapping, 6(4), 270–282. 730–737. https ://doi.org/10.1016/j.appet .2012.07.005. 1 3 Psychological Research Farrell, M. J., Zamarripa, F., Shade, R., Phillips, P. A., McKinley, M., Pardo, J., Pardo, P. J., Janer, K. W., & Raichle, M. E. (1990). The ante- Fox, P. T., … Egan, G. F. (2008). Effect of aging on regional rior cingulate cortex mediates processing selection in the Stroop cerebral blood flow responses associated with osmotic thirst and attentional conflict paradigm. Proceedings of the National Acad- its satiation by water drinking: A PET study. Proceedings of the emy of Sciences, 87(1), 256-259. National Academy of Sciences of the United States of America, Rogers, P. J., Kainth, A., & Smit, H. J. (2001). A drink of water can 105(1), 382–387. https ://doi.org/10.1073/pnas.07105 72105 . improve or impair mental performance depending on small dif- Fong, G. T., Krantz, D. H., & Nisbett, R. E. (1986). The effects of ferences in thirst. Appetite, 36(1), 57–58. statistical training on thinking about everyday problems. Cogni- Sahakian, B. J., & Owen, aM. (1992). Computerized assessment in neu- tive Psychology, 18(3), 253–292. https ://doi.org/10.1016/0010- ropsychiatry using CANTAB: discussion paper. The Royal Society 0285(86)90001 -0. of Medicine, 85(July), 399–402. https ://doi.org/10.1177/01410 Francesconi, R. P., Sawka, M. N., & Pandolf, K. B. (1984). Hypohydra-76892 08500 711. tion and acclimation: effects on hormone responses to exercise/ Sharma, V. M., Sridharan, K., Pichan, G., & Panwar, M. R. (1986). heat stress. Aviat Space Environ Med, 55, 365– 369. Influence of heat-stress induced dehydration on mental functions. Frederick, S. (2005). Cognitive {reflection} and {decision} {making}. Ergonomics, 29(6), 791–799. Journal of Economic Perspectives, 19(4), 24–42. Shirreffs, S. M., Merson, S. J., Fraser, S. M., & Archer, D. T. (2004). Gigerenzer, G., & Goldstein, D. G. (1996). Reasoning the fast and The effects of fluid restriction on hydration status and subjective frugal way: Models of bounded rationality. Psychological Review, feelings in man. British Journal of Nutrition, 91(06), 951. 103(4), 650–669. https://doi.or g/10.1093/acprof:oso/97801 99744 Sloman, S. A. (1996). The empirical case for two systems of rea- 282.003.0002. soning. Psychological Bulletin, 119(1), 3–22. h t t p s : / / d o i . Gopinathan, P. M., Pichan, G., & Sharma, V. M. (1988). Role of dehy- org/10.1037/0033-2909.119.1.3. dration in heat stress-induced variations in mental performance. Stanovich, K. E., & West, R. F. (2000). Individual differences in rea- Archives of Environmental Health, 43(1), 15–17. https ://doi. soning: implications for the rationality debate? The Behavioral org/10.1017/CBO97 81107 41532 4.004. and Brain Sciences, 23(5), 645–726. Greendale, G. A., Kritz-Silverstein, D., Seeman, T., Barrett-Connor, Toplak, M. E., West, R. F., & Stanovich, K. E. (2011). The cognitive E. (2000). Higher basal cortisol predicts verbal memory loss in reflection test as a predictor of performance on heuristics-and- postmenopausal women: rancho bernardo study. Journal of the biases tasks. Memory & Cognition, 39(7), 1275–1289. https://doi. American Geriatrics Society, 48(12), 1655–1658.org/10.3758/s1342 1-011-0104-1. Kahneman, D. (2003). A perspective on judgment and choice: Mapping Toplak, M. E., West, R. F., & Stanovich, K. E. (2014). Assessing bounded rationality. The American Psychologist, 58(9), 697–720. miserly information processing: An expansion of the cognitive https ://doi.org/10.1037/0003-066X.58.9.697. reflection test. Thinking & Reasoning, 20(2), 147–168. https :// Kahneman, D. (2011). Thinking, fast and slow. Toronto: Doubleday doi.org/10.1080/13546 783.2013.84472 9. Canada. Tversky, A., & Kahneman, D. (1974). Judgment under uncertainty: Khng, K. H., & Lee, K. (2014). The relationship between stroop and Heuristics and biases. Science, 185(4157), 1124–1131. https :// stop-signal measures of inhibition in adolescents: Inu fl ences from doi.org/10.1126/scien ce.185.4157.1124. variations in context and measure estimation. PLoS One, 9(7), Tversky, A., & Kahneman, D. (1983). Extensional versus intui- e101356. https ://doi.org/10.1371/journ al.pone.01013 56. tive reasoning: The conjunction fallacy in probability judg- Kirschbaum, C., Wolf, O. T., May, M., Wippich, W., & Hellham- ment. Psychological Review, 90(4), 293–315. https ://doi. mer, D.. H (1996). Stress- and treatment-induced elevations of org/10.1037/0033-295X.90.4.293. cortisol levels associated with impaired declarative memory in healthy adults. Life Sciences, 58(17), 1475–1483. https ://doi. Publisher’s Note Springer Nature remains neutral with regard to org/10.1016/0024-3205(96)00118 -X. jurisdictional claims in published maps and institutional affiliations. Miyake, A., Friedman, N. P., Emerson, M. J., Witzki, A. H., Howerter, A., & Wager, T. D. (2000). The unity and diversity of execu- tive functions and their contributions to complex “Frontal Lobe” Tasks: A latent variable analysis. Cognitive Psychology, 41(1), 49–100. https ://doi.org/10.1006/cogp.1999.0734. 1 3

Journal

Psychological ResearchSpringer Journals

Published: Jan 21, 2019

There are no references for this article.