In this study, 112 participants were recruited using a non-random sampling method (convenience sampling) and divided into four groups: a control group (T) and three experimental groups (E1, E2 and E3). Participants in groups E1, E2 and E3 were asked to drink quantities of alcohol to respectively approximate blood alcohol contents of 0.2 g/l, 0.5 g/l and 0.8 g/l.

Each group (control and experimental) was subdivided into two subgroups or categories: occasional consumers (OC) and regular consumers (RC). Each group comprised 28 individuals (14 occasional drinkers and 14 regular drinkers).

The Occasional Alcohol Consumer is a person whose consumption is ≤4 glasses/month and no more than 4 standard glasses on a single occasion and the Regular Consumer has a consumption ≥ 4 glasses/week) [10] .

All the participants were male. They had a mean age of 21.9 ± 3.8 years (ranging from 18 to 24 years) and a mean mass of 61.8 ± 6.1 kilograms. The inclusion criteria were as follows: male, in good physical and mental health, had at least consumed alcohol once, able to read and at least 18 years of age. People with alcohol hypersensitivity, metabolic or cognitive disorders and colour-blind participants were excluded.

The technical equipment consisted of: pure ethanol (99.8%); distilled water; a flavouring to improve the palatability of the solution; a graduated cylinder; a weighing scale; a laptop computer; a drinking glass and a four-disc test of the Hanoi tower (see Figure 1 ).

The alcohol taken up by participants had an ethanol concentration of 40% and was prepared by diluting 99.8% of ethyl alcohol with distilled water using the following formula:

$C_i\times V_i=C_f\times V_f$

C_i = initial concentration of alcohol

V_i = initial volume of alcohol

C_f = final concentration of alcohol

V_f = final volume of alcohol

On the day of the test, participants were sober for at least 24 hours and fasted for at least 8 hours to allow a rapid absorption of alcohol [11] [12] .

The test was carried out in a room isolated from all external influences. An informed consent of each participant was obtained prior to experiment. According to an interview, participants who claimed to have consumed 5 or more standard glasses of alcohol at least once a week were considered as regular drinkers and those who consumed 3 to 4 standard glasses at least once a month were seen as occasional drinkers ( [10] and [13] ).

The participant’s body mass was determined on arrival in the test room. The volume of drink to be taken by the participant was determined according to Widmark’s formula ( [14] Widmark, 1932):

$V=\frac{T\times K\times M}{P\times 0.8}$ From which we have $V=\frac{T\times K\times M}{P\times 0.8}$

T = Blood alcohol content

V = volume of alcohol consumed by the individual in ml

M = body mass of the individual in kg

K = the diffusion coefficient (K = 0.7 for men and K = 0.6 for women)

P = percentage of alcohol (40% = 0.4)

0.8 = density of ethanol.

Participants were given five minutes to drink the experimental alcohol solution. The test was carried out 15 minutes after drinking of the beverage.

In this study, the Tower of Hanoi test was used to assess the problem-solving performance of the different participants, according to Goldstein and Green (1995) [15] , this test is used to assess planning, which is a process involved in problem-solving. The Tower of Hanoi test consists of three poles (marked A, B and C) placed vertically and with several discs (four in this study) of decreasing size. Each disc was pierced in its centre to be placed around one or other of the three poles. The discs were initially placed around stake A in decreasing order of size, forming a tower. The aim of the game is to move the discs until you reach the final point where all the discs are around stake C in decreasing order. The discs may come and go freely on the stakes according to two rules:

The rest interval between two trials was one minute (1 min). This interval was occupied by distractors (a series of general culture questions).

Data collection

For this study, 4 successive trials were carried out and the data collected were the number of actions (moves) performed to solve the problem, for each trial.

Variables

The study variables were of two types:

The data collected was first entered into a spreadsheet (created using Microsoft Excel 2016), then transferred and processed using STATISTICA^® 7.1 software. This was used to generate graphs and compare the mean performance of different groups using the Kruskal-Wallis H-test (non-parametric test). To analyse performance between categories (CO and CR), the Mann-Whitney U Test (non-parametric test) was used.

The probability (p) of 0.05 was considered to be the limit of significance. Thus, if “p” is less than or equal to 0.05, the difference between compared means is significant. On the other hand, if “p” is greater than 0.05, then the difference between compared means is not significant.

The statistical meanings shown in the different figures were coded as follows: *: p ≤ 0.05; **: p ≤ 0.01; ***: p ≤ 0.001.

Figure 2 showed the number of actions performed, by group and by trial, to solve the Tower of Hanoi problem.

For trial 1, the analysis indicated H (3, 56) = 19.07; p < 0.001. Multiple comparisons revealed that the performance of individuals from group E1 (19.87 ± 2.88) was significantly better (these individuals performed fewer actions than the others to solve the problem) than that of individuals from group T (37.29 ± 21.39) (p = 0.008). Likewise, the difference in performance between group E1 (19.87 ± 2.88) and group E2 (38.07 ± 21.77) was highly significant (p = 0.007). Finally, the difference in performance between groups E1 (19.87 ± 2.88) and E3 (41.23 ± 13.41) was also highly significant. (p < 0.001). The other comparisons did not show any significant differences (p > 0.05).

For trial 2, the analysis indicated H (3, 56) = 20.66; p < 0.001. Multiple comparisons revealed that the difference in performance between group T (56.71 ± 17.14) and group E1 (26.73 ± 10.90) was highly significant (p < 0.001). Similarly, there was a significant difference in performance between individuals from group T (56.71 ± 17.14) and those in group E3 (32.31 ± 8.32) (p = 0.008). The other comparisons showed no significant differences (p > 0.05).

For trial 3, the analysis revealed H (3, 56) = 11.85; p = 0.007. This indicated that at least two groups differ significantly. Multiple comparisons revealed a significant difference between the performance of individuals from groups E1 (23.40 ± 10.37) and E2 (53.79 ± 55.78) (p = 0.006). The other comparisons did not exhibit significant results (p > 0.05).

For trial 4, the analysis indicated that the differences in performance between groups were not significant (H (3, 56) = 6.99; p = 0.072).

Figure 3 showed the number of actions carried out, by group and by trial, to solve the Tower of Hanoi problem, for occasional alcohol consumers.

For trial 1, analysis of data indicated H (3, 30) = 12.82; p = 0.005. Multiple comparisons revealed that the mean performance of individuals from group E3 (55.33 ± 10.17) was significantly higher than that of individuals from group E1 (20.78 ± 3.07) (p = 0.003). The other comparisons did not show any significant differences (p > 0.05).

For trial 2, the non-parametric analysis of variance showed (H (3, 30) = 15.42; p = 0.001). Multiple comparisons concluded that there was a highly significant difference in performance between individuals from group T (55.75 ± 16.20) and those from group E1 (22.56 ± 8.82) (p < 0.001). The other comparisons did not show significant results (p > 0.05).

For trial 3, the non-parametric analysis of variance gave H (3, 30) = 12.88; p =0.004. Multiple comparisons revealed a significant difference in performance between individuals in group E1 (19.6 ± 6.22) and those from group E2 (66.71 ± 79.38) (p = 0.045). Similarly, groups E1 (19.6 ± 6.22) and E3 (39.00 ± 5.93) differed significantly (p = 0.007). As for the other comparisons, the differences observed were not significant (p > 0.05).

For trial 4, the analysis indicated H (3, 30) = 6.71; p = 0.021. The performance of group E1 (21.00 ± 11.49) differed significantly from that of group E2 (42.71 ± 35.39) (p = 0.027). The other differences were not significant (p > 0.05).

Figure 4 showed the number of actions carried out, by group and by trial, to solve the Tower of Hanoi problem, for regular consumers.

For trial 1, the analysis revealed that at least two groups differed significantly (H (3, 26) = 17.53; p < 0.001). Multiple comparisons showed that the performance of group T (55.50 ± 21.36) differed very significantly from that of group E1 (18.50 ± 2.07) (p < 0.001). Similarly, the difference between groups E1 (18.50 ± 2.07) and E2 (40.86 ± 18.86) was significant (p = 0.012). The other comparisons showed non-significant results (p > 0.05).

For trial 2, H (3, 26) = 7.17; p = 0.066, so the groups were statistically identical in terms of performance.

For trial 3, differences in performance between groups were not significant (H (3, 26) = 3.91; p = 0.271).

For trial 4, the analysis showed that the groups were significantly different in terms of performance (H (3, 26) = 8.77; p = 0.032). Multiple comparisons revealed that there was a significant difference between groups T (49.00 ± 23.80) and E2 (22.14 ± 8.19) (p = 0.028). The other comparisons showed no significant differences (p > 0.05).

Figure 5 showed the number of actions performed to solve the Hanoi Tower problem for the CO and CR categories, for groups T, E1, E2 and E3.

For group T, the differences between CO and CR were significant on trials 1 (p = 0.001) and 4 (p = 0.013). For the other trials, the result was p = 1.00 for trial 2 and p = 0.108 for trial 3. The latter was not significant.

For group E1, the analysis revealed a significant difference between CO and CR in trial 2 (p = 0.036). For the other trials, the differences were not significant (p > 0.05).

For group E2, the analyses did not reveal any significant differences (p > 0.05).

For group E3, the analyses revealed a significant difference between CO and CR for trial 1 (p = 0.001). However, the other trials did not show any significant differences (p > 0.05).