Edited by:
Reviewed by:
*Correspondence:
This article was submitted to Personality and Social Psychology, a section of the journal Frontiers in Psychology
This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
People help each other less often and less quickly when bystanders are present. In this paper, we propose that alcohol consumption could attenuate or reverse this so-called bystander effect. Alcohol impairs people cognitively and perceptually, leading them to think less about the presence of others and behave less inhibited. Moreover, alcohol makes people more prone to see the benefits of helping and not the costs. To provide an initial test of these lines of reasoning, we invited visitors of bars in Amsterdam to join our study at a secluded spot at the bar. We manipulated bystander presence, and at the end of the study, we measured alcohol consumption. When participants took their seats, the experimenter dropped some items. We measured how many items were picked up and how quickly participants engaged in helping. Results revealed that alcohol did not influence the bystander effect in terms of the amount of help given. But importantly, it did influence the bystander effect in terms of response times: people who consumed alcohol actually came to aid
Imagine having dinner with three colleagues in a nice, but crowded restaurant. Suddenly you hear some loud coughing sounds. The sounds seem to come from a man who sits alone at a table not too far from yours. Slowly his face turns red, and you suspect he is choking. Would you leave your colleagues and help this man? From what we know about the bystander effect, especially when other people are around, helping is unlikely in this situation, or it may take a long time before someone helps. In the present study, however, we examine the possibility that if the dinner was accompanied by some glasses of wine, people may in fact be quite ready to intervene, even when other bystanders are around.
Alcohol consumption is often seen as one of the most important determinants of anti-social behaviors such as vandalism and interpersonal violence (
Since the early work of
The bystander effect became publicly known through news reports about bystanders of violent crimes who neglected to intervene sufficiently or timely. For instance,
The decision to help or not is often based on an implicit calculation in which people outweigh the (emotional) costs and benefits of helping, versus those of not helping (
First, a well-known process is referred to as
Audience inhibition can happen for at least two reasons. First, people fear they misinterpret the situation and help is actually not required. The choking man in the restaurant could just be coughing and looking red because he has a fever. Indeed, in a study by
Thus, across a number of situations, the presence of bystanders exerts inhibitory effects on helping. Importantly, inhibition not only undermines helping, it also makes people respond more slowly—which is important, especially as in most emergency situations time matters greatly (e.g.,
Conventional wisdom holds that alcohol consumption is associated with multiple health and social concerns (
One of the main reasons alcohol changes social behavior is because it impairs cognitive functioning which is needed to inhibit certain social responses (e.g., intoxicated people are more likely to engage in self-disclosure, and sometimes aggression, than are non-intoxicated people (
Alcohol myopia is related to the process of cognitive impairment. It entails that alcohol consumption makes people more attentive to salient cues in the environment, while it decreases attention to less salient cues (
Alcohol consumption can even change bystander behavior without actually affecting their cognitive functioning. In general, people expect to behave less inhibited when they are drunk. The mere belief of people that they are intoxicated still yields uninhibited behaviors, even when people did not drink alcohol at all (
Based on a cost-benefit perspective of helping (
The present study was designed to provide an initial test of the notion that alcohol can diminish the negative influence of bystander presence on helping. In our pursuit of ecological validity, we faced several challenges. The most important challenge was that we wanted to provide a test in a real world, rather than the lab, as drinking in the lab may be quite a different experience from drinking in a more natural, social setting such as a bar. The study, therefore, takes place in actual bars in Amsterdam. We empirically test the notion that alcohol can diminish the negative influence of bystander presence on helping. We test this by measuring the proportion of helping and the onset time of helping during an item dropping paradigm in a bar (see also,
In line with the bystander effect, we expect to find that the presence of bystanders decreases the amount of help given, and that it increases the time people need to come to a decision to help. However, our focus lies mostly on the interaction effect of alcohol and bystander presence. Alcohol is known to influence cognitive functioning (“The wine made me less reflective”;
A total of 120 people (59 female, 61 male; mean age = 31.22,
Upon entering the bar, visitors were asked to sign an informed consent stating that they were willing to participate in a study later that evening. After at least 20 min, participants who agreed were asked to sit at a table at an isolated location inside the bar, where the experiment would take place. At this table, there were either two confederates (male) ostensibly filling out questionnaires, or there was no one at all. The two confederates sat next to each other, leaving two spots open on the other side of the table. The experimenter (female) sat next to the participant to explain what the questionnaire was about and that when the study was done, the participant would be asked to blow in the breathalyzer.
While the experimenter handed over the questionnaire, she deliberately knocked over a canister of 20 mouthpieces for the breathalyzer. The experimenter made sure every mouthpiece would fall out and roll over the floor. She would then hand over the questionnaire and slowly started picking up the mouthpieces at a rate of about one per 5 s, starting with the farthest first. At the time the canister was knocked over, an observer across the room, would start a stopwatch, and note down the time it takes the participant to engage in helping, and how long the participant would help. In the bystander condition, the bystanders were instructed to look up when the “accident” occurred, but they then continued filling out the questionnaire.
After the participants chose to help or not, and they finished the questionnaire, they were asked to blow in the breathalyzer. The BAC value was noted, the participant thanked and thoroughly debriefed.
Helping behavior was measured two times. First, people were coded as helpers if picked up mouthpieces, or as non-helpers if they focused only on finishing their questionnaire. Second, helping behavior was quantified by counting the number of mouthpieces the participant picked up. This could range from 0 to 20 and was normally distributed (kurtosis = -0.46, skewness = -0.04).
As soon as the experimenter dropped the jar of mouthpieces, a confederate to the experimenter in an adjacent room would start a timer on a stopwatch. The timer was stopped as soon as the participant picked up the first mouthpiece. There was no response time for people who did not help. Response times were log transformed to more closely fit normality assumptions (kurtosis = -0.38, skewness = 0.22 after transformation).
We measured drinking behavior in four ways: we measured BAC by means of a breathalyzer, but also asked how many alcoholic drinks they had during the last 3 h. Furthermore, we asked participants two questions, which they could answer on a seven-point Likert scale: “My current state is” with the answer anchors, 1 =
In line with research on the bystander effect we expected that the presence of bystanders diminishes helping. First, by the likelihood of people helping the experimenter or not, and second, by the number of mouthpieces picked up. We, however, also hypothesized that for people under the influence of alcohol, the bystander effect diminishes or even reverses.
First, we used binary logistic regression to test the effect of bystander presence (coded 0 for absent, 1 for present), alcohol consumption (centered) and their interaction term on the decision to help or not. The analysis showed that in the presence of others participants were (marginally) less likely to help (57.6% helped) than when they were alone (73.8%), Wald’s χ2(1,
Helping behavior as a function of alcohol consumption and bystander presence.
Response time. | Items picked up. | Helping yes/no. | |||||||
. | . | Exp b. | |||||||
Constant | 0.51 | 0.42 | 11.38 | 0.54 | -0.31 | 0.26 | |||
Alcohol consumption | 0.28 | 0.19 | 0.19 | 1.38 | 2.49 | 0.07 | -0.42 | 1.46 | 16.74 |
Bystander presence | -0.02 | 0.06 | -0.04 | -3.19*** | 0.83 | -0.41 | -0.74 | 0.39 | 1.04 |
Alcohol consumption Bystander presence | -0.86* | 0.33 | -0.36 | 0.24 | 4.19 | 0.01 | -1.92 | 2.15 | 10.01 |
Second, from those participants who actually helped (32 females, 47 males), we regressed the number of mouthpieces they picked up, on bystander presence (0 for not present, 1 for present), alcohol consumption (centered), and their interaction term. Consistent with the bystander effect, we found that the presence of bystanders diminishes the number of pieces participants picked up,
In line with research on the bystander effect, we expected that it would take people longer to engage in helping when others are present. However, we expected that alcohol consumption attenuates or reverses this bystander effect. We regressed reaction time (in seconds) on bystander presence (0 for not present, 1 for present), alcohol consumption (centered), and their interaction term (see
As expected, we found a significant interaction effect of bystander presence and alcohol consumption
Taken together, the results reveal that alcohol consumption indeed changes the classic bystander effect, in terms of reaction times. Alcohol consumption does not simply attenuate the influence of bystander presence on the decision making process, people who consumed alcohol actually became
In the introduction, we outlined three key processes that help understand how alcohol consumption could overcome the problem of bystanders intervening too slowly or not at all, namely disinhibition, selective attention, and behavioral expectations. In the current study, one of the first that examines the role of alcohol in bystander behavior, it was impossible to completely disentangle these potential explanations. However, the outcomes did help us to rule out some alternative hypotheses. For instance, based on the disinhibition hypotheses, one could have alternatively reasoned that alcohol lifts inhibitions toward selfish responses, such as not helping. Future research could aim to better disentangle the different processes we have outlined and perhaps even other, related, processes such as expectancies of self-confidence and heightened power after drinking. Two related processes which are shown to influence bystander intervention (e.g.,
The foremost finding from the current contribution is that alcohol does not simply attenuate the entire bystander effect, but only increases
We would like to note some strengths and limitations of the current work. Especially when making the step from the lab to the field, one must recognize the power of naturalistic influences. Small bars can be very noisy, both in a literal and methodological way. By inviting the participants to come to a secluded spot, we aimed to avoid such noises, but were not always completely successful. Another important limitation is that alcohol was only measured, not manipulated. Because of this limitation we must remain cautious about interpreting the results in terms of causality. At the same time, this is also adds to the major strength of the study, namely ecological validity. The experiment was conducted in a setting where the bystander effect may be very pronounced, and involved people who are often exposed to bystander situations, as bars are a common setting where people need help (e.g., accidents due to crowdedness, sickness due to alcohol intoxication, and public violence;(
The bystander effect has since the early studies (
MvB and J-WvP designed the experiment, MvB Supervised data collection and prepared the first draft of the manuscript. All authors listed have made substantial, direct and intellectual contributions to the work by discussing the results and implications, and by commenting on and editing the manuscript at all stages.
The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
Although not hypothesized, we repeated each key analysis with participants’ sex and the resulting interaction terms as additional variables, in order to test if the effects were different for males and females. We, however, found no significant interaction effect for sex. Helping behavior: Wald’s χ2(1,