A special thanks to the Australian College of Applied Psychology for performing the research and providing the results that contribute to the below information.
The term “Hive Mind” was first used in Olaf Stapledon’s science fiction novel ‘Last and First Men’ (1930). It was a term to describe a group mind in which the linked individuals have no identity or free will and are mind controlled as extensions of the hive mind. This term is no longer fictitious and used in modern-day research (Thompson F., 2016; Haidt, Seder, & Kesebir, 2008; Gaggioli, 2017; Aron, 2012).
Historically, group intelligence was used by societies and cultures for survival. It allowed for an awareness of geographical locations, sharing of knowledge and protection of communities from harm. Cultures develop how individuals perceive, comprehend and interpret the world around them. Individualistic cultures are likely to have an independent view of themselves, whereas collectivist cultures tend to have an interdependent view of themselves (Markus & Kitayama, 1991).
Although societies still use cultural knowledge to subsist and remain aware, communications have evolved beyond word of mouth. Social media plays a pivotal role in delivering news (Fletcher & Nielsen, 2018), politics (Kruse, Norris & Flinchum, 2018), public service announcements (DePaula, Dincelli & Harrison, 2018), health and science (Lee, VanDyke & Cummins, 2018; Jarreau & Porter, 2018). Research performed indicates two-third of American adults get news on social media sites. The platforms where users were primarily exposed to news is Reddit, Twitter and Facebook (Matsa & Shearer, 2018).
While social media is key in delivering information; controversial individuals and social media bots work to spread false information on a large scale (Kumar & Shah, 2018). In a study of data collected from 2006-2017, over 126,000 stories shared by 3 million international users was discovered to be false news and information. The study proved that a false story reaches 1,500 people six times quicker on average than a true story does (Vosoughi, Roy, & Aral, 2018).
The concept of false information has existed historically in both individualistic and collectivist cultures. Misperceptions are prevalent in many ongoing debates in politics, health and science (Flynn, Nyhan, & Reifler, 2017). Debates surrounding vaccination and autism gained momentum in the 1990’s (Wakefield, et al., 1998). Despite various research over the years proving the opposite (Madsen, et al., 2002; Plotkin, Gerber, & Offit, 2009; Hviid, Hansen, Frisch, & Melbye, 2019), social groups and movements exist in favour of linking vaccinations to autism. Theories of reptilian humanoids controlling politics are an obvious falsehood that over 12 million Americans consider factual according to a poll conducted by Public Policy Polling (Williams, 2013). In some collectivist cultures such as Asian cultures, there is a strong belief in numbers (Tse, 2015), whilst African cultures hold a strong belief in witchcraft (Geschiere, 2008).
There are many reasons that lead to conformity. Disentangling truth from falsehoods when mixed from different sources make the challenge of distinguishing the truth even more difficult (Rapp, 2016). Informational influence contributes when a situation is ambiguous (Morton & Gerard, 1955) and when information is reiterated, it causes an illusory truth effect (Hasher, Goldstein, & Toppino, 1977). With the ability of resharing information on social media, it is inevitable that users will be presented with reiterations of information.
The statistics of social media platforms indicate a mix of both collective and individualist cultures, which support evidence that social media is not targeted at one culture. For example; the largest population of users on the social media platform Twitter, were from the United States (47.05 million), followed by Japan (38.6 million), United Kingdom (13.6 million) and Saudi Arabia (11.27 million) (Statista, 2019). On Facebook, the leading figures are represented by India (300 million) and followed by United States (210 million), Brazil and Indonesia (130 million) (Statista, 2019).
The aim of this study was to establish conformity in cultures. The original Solomon Asch conformity test was a social experiment where a group of participants would partake in a perceptual task of verbally identifying a line followed by 3 other lines labelled A, B and C. Participants included one non-actor and remaining participants were actors instructed to respond correctly but on some trials to unanimously respond incorrectly. The post interview of subjects revealed conformity in answers to avoid standing out, despite their accurate judgement (Asch, 1956).
This study was modified to suit modern research of conformity. Students participated in an online version of the experiment and were presented on some trials with the actor’s responses. The hypothesis was that participants could identify the right answer but when placed in no match trials, participants mainly belonging to the collectivist culture, will conform to confederate answers.
Participants were 224 (85 male and 139 female) undergraduate students from an Australian tertiary education institution. All participants were required to be over 18 years of age and participated in return for credit as part of a unit of study. Failure to participate did not impact the participants’ progression in the unit of study. The mean age of the participants was 27.62 years (SD = 3.38); male mean age was 28.52 years (SD = 2.14), and female mean age was 26.85 years (SD = 3.86).
The conformity task used in this experiment was modelled on the original work of Asch (1956). Here participants were presented with a target line to the left of the screen and asked to select one of three test lines located on the right of the screen that most closely matched the target line on the left. Participants indicated their choice amongst the three lines by selecting the corresponding letter key on the computer keyboard.
On half of the trials presented (n=48), participants made their judgement based solely on the line stimuli presented (SELF trials), while on the other half of the trials participants were able to view the responses of five confederate responders listed to the far right of the screen (CONFEDERATE trials). See Figure 1 for a depiction of the two trial conditions. The CONFEDERATE trials were designed to mimic the Asch (1956) condition where the participant was required to respond to the stimuli in the presence of five in-person confederates. In the current study the confederates were not in-person responders, but rather responses pre-selected by the experimenters and presented to all participants on the same trials. To begin, all CONFEDERATE trials included responses that accurately indicated the matching test line. Following the first seven trials the remaining trials were divided between accurate (35%) and inaccurate (65%) responses.
Figure 1. Depiction of the two trial stimulus conditions (a) SELF trials where the participant is provide with no information about other responders selections and (b) CONFEDERATE trials where the responses of five confederates are provided.
Participants were seated before a computer screen at an approximate distance of 40-50cm. Participants responded throughout the task using the computer keyboard by pressing the keys [A], [B] and [C] to indicate their response on the experimental trials. Participants were required to log onto a secure server, Inquisit, which hosted the entire experiment. The experiment was divided into two stages; demographic questionnaire which collected information about participant gender, age, cultural background, level of education and household income (Stage 1), and the conformity task (Stage 2). Participation took approximately 25 minutes.
Prior to running the statistical analyses all data were screened for normality. As a result of missing data, the responses of 205 participants were retained for further analysis. Following the removal of the data of one participant, identified as a univariate outlier, the data was once again screened for normality and homogeneity of variance with no violations present.
The data was collapsed into two groups, collectivist and individualist, according to the work of Hofstede (1980) and a t test was conducted to determine whether a difference existed in the level of conformity between the two cultural groups. Conformity was determined by comparing the number of incorrect responses selected by a participant when completing SELF compared to CONFEDERATE trials.
Initial analyses revealed that in general, levels of conformity in this experiment were quite low, that is, when a participant made an error in line selection the error was not necessarily unique to responses made on CONFEDERATE trials; that is, participants made errors on trials where their decisions were based entirely on their own judgement and not impacted by confederates. On those trials where errors were made on CONFEDERATE trials approximately 35% of the errors were in agreement with the confederate responses, while 65% of errors were different to the confederate responses. Interestingly, on appraisal of trials where the Target Line did not match with any of the Match Lines (No Match Trials), the responses of the confederate selection did appear to be of greater influence. Specifically, on No Match Trials 81% of participant responses matched that of the confederates.
An analysis of the influence of culture on the conformity exhibited by participants was significant. As can be seen in Figure 2, individuals belonging to collectivist cultures tended to make more errors on CONFEDERATE trials than did individuals from individualist cultures. Importantly, the incorrect decisions made by individuals belonging to a collectivist culture matched the selection of the confederates on the trial. That is, these individuals did not merely make the wrong decision but instead that wrong decision was in line with the confederate majority. This suggests that individuals from individualist cultures are less impacted by the implied presence of others (n=34% of CONFEDERATE trials) and are more likely to oppose an opinion held by the (implied) majority than individuals from collectivist cultures (n=55% of CONFEDERATE trials).
Figure 2. The mean number of incorrect responses (in%) made in the CONFEDERATE condition for individuals from collectivist and individualist cultures where the response was consistent with the confederate line selection.
Based on a research by Roy Morgan (2018), Australians spend 21.9 billion hours a year on the internet. The conditions to perform this experiment suited the current climate where users input their answers on a web-based application using technological devices (computer screen and keyboard). The experiment was modelled on the original work of Asch conformity where participants performed the experiment on computer screens and were instructed to select one of three lines on the right of the screen by pressing keys A, B and C to identify the line closely matching the target line. The results support the hypothesis that individuals can identify the difference between a false presentation and the truth, but when confronted with uncertainty, conformity will be present.
The statistics of collectivist cultures superseding individualists on incorrect response of confederate trials was consistent with previous research (Hofstede, 1980; Markus & Kitayama, 1991; Oh, 2013) at an individualist’s scoring n=34% of CONFEDERATE trials against individuals in collectivist cultures at n=55% of CONFEDERATE trials.
Another critical finding related to the hypothesis was individuals’ levels of conformity in this experiment were quite low when responses did not reveal confederate scores. When presented with no match trials, 81% of responses matched the confederate response. The findings recognise the presence of informational influence when confronted in ambiguous situations, is consistent with previous research (Bishop & Myers, 1974; Bikhchandani, Hirshleifer, & Welch, 1992). The uncertainty faced while making these decisions are referred to as doubt. While this may be a more cognitive approach, doubt is classed as a negative emotional response. Emotional responses elicit a stronger belief and reaction towards a response (Vosoughi, Roy, & Aral, 2018). Combined with information reiterated five times by all confederates leading to illusory truth (Fazio, Payne, Brashier, & Marsh, 2015) aided respondents with their answers.
Females tend to conform more than males (Eagly & Chrvala, 1986; Santee & Jackson, 1982) and an observation made on the participant’s gender was a ratio of 85 male to 139 females, which may have impacted the conformity levels. Future researches could have an even ratio of male to female participants for a uniform research sample.
Other suggestions for future research would be to draw attention to the confederate answers. This could be done by:
- Provide a dynamic presentation of confederate information on the screen rather than static presentation of confederate’s names and answers when participants are introduced to each new trial. By sensing a movement on the screen, participants lose focus on responding to the right answer and like the traditional test, would feel a pressure to conform.
- A restriction on screen resolution size. The web application was designed to stretch across the screen. Since participants focused on the target and match lines to provide answers, if a screen resolution were large enough to spread, it would be easy for participants to ignore confederate responses. Web-based testing programs can impact verbal test scores. (Bridgeman, Lennon, & Jackenthal, 2001)
- Eye Tracking technology to monitor the frequency of participants referring to confederate answers. This would allow measurements of how often individualist and collectivist cultures relied on confederate responses and/or how often they paid attention to responses.
Suggestions to improve future tests would be to provide participants with a post-interview questionnaire in the future that includes the below:
- Were participants aware of the actor’s responses being correct at certain times or did they only notice incorrect responses?
If the response is NO, this question will identify if individuals resisted reliance on confederate responses during no match trials as it was established that confederates were unreliable.
- When provided with no match trials, did participants rely on responses from confederates?
If YES is selected, a further prompt inquiring why they did rely on responses. This would match with the traditional Asch Conformity post-experiment interview where participants indicated that they conformed to fit in with responses from other candidates.
- A selection of emotions when provided with no match trials and a text box to provide reasoning for their choice.
This would allow to verify if emotions played a part in their responses. The Emotion Regulation Skills Questionnaire (ERSQ) (Berking& Znoj, 2008), will allow participants to specify the effort taken, stress, difficulty, distress and challenge levels.
As we have evolved, we no longer only communicate within our social groups. Individuals from all cultures connect on a global level to collaborate on united goals. This research has proven that regardless of the method we choose to connect to the hive mind, the cultures an individual identify with can shape conformity.