Human brains are quite inefficient cognitive machines. All of us regularly fail to see things accurately around us, we make poor assumptions, we misremember facts, we give in to cognitive biases, and we often make decisions based on emotional whims, rather than thinking rationally. In short, as individuals, we get things wrong, a lot of the time!
So, do groups fare better? Are two (or more) heads usually better than one at problem-solving and decision-making?
It might perhaps seem obvious that groups, by pooling information and experience of a number of people and engaging in constructive argument, will naturally tend to do better. Equally, it’s easy to appreciate how a group can help to mitigate against common individual biases and weaknesses (e.g. inaccurate rating of one’s abilities, over reliance on one’s own experience, and preferring options that favour one’s own interests).
Well, actually, groups can be better than individuals – but not always. In fact, groups come with many of their own issues and problems and, in many circumstances, can actually perform worse than individuals. Such problems are not widely understood by leaders, not least because many of the biases and processes behind them operate subconsiously.
So, leaders need to be aware, firstly, of some of the key problems groups can face and, secondly, be familiar with a range of measures they can take to try and address those problems.
Problems that leaders can face with groups
Here’s a summary of some of those problems that frequently affect groups:
i) Amplifying of indvidual biases: Whilst groups can sometimes usefully moderate certain individual biases (including, research has indicated, individuals over-stressing their own experience or beliefs, prioritising more memorable or recent facts, and ‘anchoring’ biases), conversely, risk of other biases can increase in groups – especially excessive optimism/overconfidence, belief in planning, commitment to a course of action that is already failing, and stereotyping..
ii) Group polarisation: In groups – especially closely-knit groups – deliberating together usually reduces variance and most members come into accord with one another and also tend to become much more confident of their collective judgment (whether it is actually correct or not). Unfortunately, this coming together of minds can often go further and members’ initial views or preferences become more extreme and more strongly felt (due to the social reinforcement effect of members seeing others sharing the same views as theirs). This makes the whole group become more obstinate, defensive and closed-minded – particularly vis-a-vis goup ‘outsiders’ who hold contrary views.
iii) Lack of diversity: Where members of a group are very similar in background, circumstances, opinions or beliefs, this can certainly increase group cohesion. However, a lack of diversity in such respects can easily hinder independent thinking by group members, limit the variety of experiences and opinions heard, and thwart wide-ranging discussion.
iv) ‘Cascading‘: This refers to the tendency sometimes in groups for members to follow (herd-like!) the statements and actions of those who speak or act first. Often this is because the members following judge that they have inferior knowledge, information or skill regarding the subject in question and/or simply don’t wish to spend energy or time thinking about the issue themselves.
v) Groupthink: This quite well-known term refers to when group members tend towards uniformity in their views because they do not want to upset the cohesiveness of the group or risk personally losing the liking or approval of other group members.
Other problems leaders can face with groups
vi) ‘Happy-talk‘: This occurs when the group leader focuses on maintaining a positive and enjoyable atmosphere in the group and discourages critical questioning, negative thinking, or challenging of ideas. This leads group members to avoid being fully honest or open and referring to any bad news or issues. The group becomes complacent and dysfunctional. Instead, the leader ideally needs, alongside being cheerful and positive, to show some degree of (moderate) anxiety and even doubt or scepticism in order to help stimulate constructive debate.
vii) Domineerig, confident or expert individuals: Where a group contains members who have a dominating personality or style, or they simply appear more self-confident in their views, then such individuals may well sway the opinions of several other members. The risk is that wider discussion and the group moves too quickly to accept the view of the confident individual, whose opinion may not actually be fully correct or the most suitable course of action.
viii) Leader’s ‘halo-effect’: Similarly, this is where group members follow the views of the leader because of his/her authority or status or, simply, perceived strength of personality. Often some group members will ‘over-perceive’ attractive qualities in the group’s leader (for example, the leader’s jokes seem funnier, or their perspective seems wiser), which casues them to go along with his/her views without much thinking of their own.
ix) Shared information bias (also known as ‘hidden-profile effect’): this well-proven problem refers to the tendency in groups for members to focus on information that is alreay known to and shared by all group members and the hesitancy sometimes of members to reveal/share information that is only known to them individually (or to a minority of members). This risk of limited disclosure can often seriously undermine the quality of group discussion.
x) Competing goals & social loafing: As humans have different goals and interests, it is natural that some members in a group may prioritise what suits them personally rather than making a good group decision. Ensuring that the group does not damage a member’s personal status or reputation is an example of a common concern. Another example is the concern sometimes of a group member to lessen personal accountability by limiting participation in the group and sinmply going along with whatever the majority views ends us as (a behaviour known as social-loading or free-riding).
How to improve the effectiveness of groups
In theory, it is easy to think how a group should be able to be more efficient than individuals – by pooling information and experience and letting the ‘give and take’ of deliberation arrive at an overall judgment or solution of the whole group which is better than any individual member’s judgment. It is also easy to think how the quality of the overall viewpoint would be especially good if the group contains some individuals who are particularly well-informed or experienced about the issue at-hand and/or the group includes members who have a useful variety of relevant skills.
But, unfortunately, because of the potential impact of the above range of problems, groups are far from being more effective decision-makers than individuals. Principally this is due to the distorting influence of the ‘social dynamics’ between group members (i.e. how they behave to each other inside the group) and, secondly, the typically limited disclosure and aggregating of information held by group members.
So, what are some useful measures leaders can take to try and reduce the effect of the above problems? Here are some specific actions I would recommend:
i) Collect information and opinions from individuals before the group meets: To ensure the collective knowledge of a group is properly accessed and to counter bias and groupthink, try gathering input from group members individually and anonymously before they meet together. Then circulate a shared document asking for everyone to comment on ideas or suggestions received, again independently and anonymously, without assigning any of the comments to particular team members.
ii) Split up the group decision-making process into two (or three) stages: Too often groups close down discussion and make a decision before they have generated enough ideas. So, try relying less on just one meeting. An effective approach can be to first hold a group meeting aimed at generating ideas and perspectives – where a focus on divergent, creative thinking is particularly useful – and then a second meeting to focus on crtiically assessing ideas and reaching a solution, where a more convergent thinking focus is appropriate. The other advantage of this split approach is that you can include some individuals in each group who are known to have a stronger affinity with the focused thinking approach needed.
iii) Use a diverse group rather than a homogenous group (most of the time): Groups whose members vary in their characteristics and have different points of view usually more effectively counter biases. But not always – context matters. When the task or decision is relatively complex or needs innovative thinking, then a diverse group will especially tend to do best. However, if the issue is relatively straighforward or needs quite structured or controlled thinking, then a homogenous group will tend to do better.
iv) Use a small group when you need to make an important decision: Research shows that groups with seven or more members are likely to make more biased decisions. By keeping your group to between three and five or six people – a size that people more naturally gravitate toward when interacting – you’ll be able to reduce the effects of bias but still benefit from multiple perspectives.
v) Ensure a safe, respectful and open atmosphere in the group: The group leader has a vital role to play in ensuring that all group members feel they can speak up without fear of retribution (what is sometimes called ‘psychological safety‘). His/her manner and facilitative approach must be encouraging, fair and positive. Discussion and comment should be centered around issues (rather than personalise any issues) and the leader should avoid giving his/her opinions on a topic until everyone else has spoken. If the leader is not a good faciltitator, engage an external, professional facilitator, which then allows the regular group leader to focus on contributing to the discussions. Also, of course, a friendly, open tone to a group meeting will be achieved more easily if all (or most) of the group members (not just the leader or facilitator) have been selected partly because they have good empathy and social-sensitivity skills.
vi) Appoint a team member to be ‘devil’s advocate’: Give one of the team the deliberate role and right to ask questions and challenge colleagues’ thinking during the discussion and decision-making process. For larger groups of seven or more members, perhaps appoint two such advocates, so that a sole dissenter is not isolated by the rest of the group.
vii) Assign specific roles to group members: To help achieve a more complete and rigorous discussion and assessment of the problem at hand, allocate to each group member a specific angle or area of concern to focus their thinking and comments around. For example, for a group discussion about assessing a potential new product, assigned roles could include the company’s marketing team, current customers, the main competitor, the company’s finance director, and the company’s production or technical director.
viii) Plan some structure to group discussions in advance: In a group meeting discussion will be so much more efficient and fruitful if the leader (or facilitator) thinks carefully in advance about what range of issues needs to be considered and defines a suitable outline framework (semi-structure) for running the meeting, including a rough time plan and set of specific questions or decisions that need to be addressed for each part of the meeting. Without some degree of structure, deliberation risks becoming unfocussed and there is more room for the behavioural dynamics of a group to skew or distort how matters are considered (e.g. domineering members influencing quieter members).
ix) In discussions, particularly discourage bias at the start and when views/votes are needed at the end: At the opening of a discussion session, it’s a good idea if the leader proactively reminds the whole group of the general risk of bias and asks for members actually to suggest any particular biases they think may apply in the upcoming discussion. And when the time comes for each individual to indicate their final view on a topic, ideally try and let each person reveal their position at the same time, independently and confidentially: in a virtual discussion, this may be quite easy because group members can use their online ‘vote’ button and/or use the private ‘chat’ facility to indicate their judgment to the group leader/facilitator.
x) Align incentives to ensure group members are motivated to work together well: Check that individuals in the group don’t have personal goals, incentives or other rewards in their jobs that are in conflict with the group’s overall aims – for example, a board director who is on a bonus scheme based on sales growth whilst another director’s bonus is based on profitability. Where incentives are not in harmony, they should be adjusted.
Consulting experts and others ‘outside’ the group
Beyond the above list of measures for improving group deliberations, there is also the opportunity sometimes for groups to consider consulting or involving other people outside the group, to access wider advice or opinion.
People who have a relevant expertise are an obvious choice. However, care should be taken. Turning to an expert for some specific piece of information or provision of a specific skill for application today can certainly serve a group well, but research shows that if you want to use an expert rather to give advice or predictions about the future, then experts frequently prove much less competent than you might expect.
Often this is simply because their knowledge and skill are founded on what they have learned or seen in the past and the unpredictability of the future does not, of course, mean that the future will be necessarily like the past! Also, research actually suggests that the better known an expert is, often the poorer his/her predictions are: this is because they are more concerned typically to protect their ‘status’ – which is based on past knowledge and beliefs – rather than being open to seeing things afresh. Sadly, though, users of experts are often ‘taken in’ by experts’ fame or apparent high level of confidence, which can bely their true competence.
So, avoid blind trust in expert opinion! Consult an expert, yes, but don’t actually make them part of your decision-making group itself, as they are very likely to sway others’ views too easily. Better still, if possible, seek the views of several experts rather than depending on just one. The average answer from a group of experts will always statistically have an advantage over individual experts.
Beyond consulting some experts, you might want to canvass the views of a still larger number of people – the so-called ‘wisdom of the crowd’. You may wish to do this principally becuase constitutional or ‘political’ factors in your organisation expect wider involvement of colleagues in decision-making, or you may think simply that consulting more peope beyond just a ‘core’ group will tend to yield a better informed, better quality decision overall. In such circumstances, of course, it is again the ‘average’ (i.e. middle or most common) viewpoint or opinion found from across everyone that will be taken to be the final decision.
But, actually, it would be a big mistake to believe that the best approach with group decision-making is always to ask a larger number of people and to take the average answer. Research indicates that this approach is likely to work only under particular circumstances: those in which many or most people are more likely than not to be right – because, crucially, they have a good level of relevant knowledge and they are not materially biased or prejudiced in some way about the issue at-hand (e.g. favouring older theories) or in their general cognitive ways (e.g. a tendency toward unrealistic optimism).
Additionally, the views of everyone should be collected separately and independently with no chance of any individual knowing any other’s viewpoint or judgment (thus avoiding any risks of social influencing in the process). And another ideal feature of large groups for decision-making is that they should have a good level of diversity (in terms of group members’ characteristics, circumstances, beliefs etc), especially if the problem or issue is complex, novel or challenging.
In truth, these conditions are usually difficult to achieve or are rarely followed to a signficant degree (for example, a potentially successful scenario could be a marketing director asking his marketing team whether to hire a certain job candidate, but it might be unreasistic to expect each team member to give their assessment of the candidate privately/anonymously). It is simply unrealistic to expect that all individuals in a large group will be free of any prejudices or biases themselves, that they all will have good/relevant knowledge, and there will be no communication or influencing between indviduals at all. Sadly, though, this leads to the fact that, if and when a large number of people is consulted on an issue, then the likely result is that the average position or judgment taken will be wrong (or at least ‘sub-optimal’).
Given this reality, therefore, in most cases, it is normally better for leaders to avoid seeking widespread views from larger numbers of people. Instead, the most effective thing to do will usually be for leaders to focus their efforts on relying on a smaller group of suitably well-qualified individuals and doing as much as possible to ensure robust and effective deliberation by that group.
In overall terms, groups can be very suitable and effective for supporting decision-making and problem-solving in organisations, offering many potential benefits over individuals working alone. However, groups do come with many potential risks and problems of their own, as explored above. So, leaders need to be familiar with some of the measures suggested above how to mitigate those challenges – with group composition and deliberation process being key areas to look at.