Psychology Group Leading Proposal Assignment
Group Dynamics
Donelson Forsyth
Forsyth, D. (2018). Group Dynamics (7th ed.). Cengage Learning US. https://mbsdirect.vitalsource.com/books/9798214344799
Chapter 10. Performance
People join with others in groups to get things done. Groups are the world’s workers, protectors, builders, decision makers, problem solvers, and idea generators. From the simplest of situations where each member works on a separate task with others nearby to more complex ones requiring prolonged collaboration, groups can accomplish goals that would overwhelm lone individuals. Not every group reaches its full potential, but when highly motivated members work well together they can master the most difficult of problems—even ones that require creative, novel solutions. Creativity comes naturally to groups.
When and why does working in the presence of other people facilitate performance?
When do people give their all when working in a group?
When do groups outperform individuals?
What steps can be taken to encourage creativity in groups?
Chapter Outline
Social Facilitation
Performance in the Presence of Others
Why Does Social Facilitation Occur?
Conclusions and Applications
Social Loafing
The Ringelmann Effect
Causes of and Cures for Social Loafing
The Collective Effort Model
Working in Groups
The Process Model of Group Performance
Additive Tasks
Compensatory Tasks
Disjunctive Tasks
Conjunctive Tasks
Discretionary Tasks
Process Gains in Groups
Group Creativity
Brainstorming
Improving Brainstorming
Alternatives to Brainstorming
Chapter Review
Resources
The Miracle on the Hudson: Working with Others in Groups
US Airways Flight 1549 from New York’s LaGuardia Airport to Charlotte, North Carolina, began uneventfully. The flight attendants helped the 150 passengers settle into their seats. The baggage handlers packed the cargo hold of the Airbus A320 with luggage and sealed the doors. Nothing was amiss as the two-man crew piloting the aircraft—Captain Chesley B. “Sully” Sullenberger and first officer Jeffrey B. Skiles—completed the preflight checklist. Flight 1549, with Skiles at the controls, lifted off from Runway 4, and the local controller passed the flight to Patrick Harten, the air traffic controller for LaGuardia departures. The time was 3:26 PM.
Less than five minutes later, the Airbus A320 was sinking in the freezing waters of the Hudson River. The cause: The aircraft, during its ascent from take-off, collided with a flock of geese, severely damaging both engines. The A320 was designed to be flyable with a single engine functioning, but the loss of two engines was catastrophic. Sully, who was more experienced flying the A320, immediately started the auxiliary power unit before telling Skiles, “My aircraft.” Skiles responded, “Your aircraft,” before turning his attention to restarting the damaged engines (Brazy, 2009, p. 38).
Sully radioed ground controller Harten: “Mayday. Mayday. Mayday. This is, uh, Cactus fifteen thirty-nine. Hit birds. We’ve lost thrust in both engines. We’re turning back towards LaGuardia” (Sullenberger, 2009, p. 217). Harten initiated emergency procedures, including communicating the problem to airport and municipal emergency response personnel. The engines would not restart, however, and the plane—now gliding without any engine power—did not have enough altitude or speed to reach LaGuardia. Sully considered taking the plane to an airport he could see in the distance (Teterboro Airport), but decided he could not reach that airport, either. He radioed Harten, “We’re gonna be in the Hudson.” Harten’s response: “I’m sorry, say again, Cactus?” (Brazy, 2009, p. 39).
New Yorkers watched as the A320, traveling at about 150 MPH, splashed down in the middle of the Hudson River, just south of the George Washington Bridge. The ditching damaged the underside of the aircraft, which began sinking as water flooded the rear of the passenger cabin. Sully gave the order to evacuate, and the passengers exited as quickly as they could onto the wings and an emergency slide at the front exit. Within minutes, a flotilla of rescue vessels assembled around the aircraft, and the passengers and crew were pulled from the icy water to safety. Some required hospitalization, but not a single life was lost. The media dubbed the event the “Miracle on the Hudson,” and the crew members were awarded the Master’s Medal of the Guild of Air Pilots and Air Navigators for their skillful handling of this emergency.
Flight 1549 is a study of groups that get things done. The three flight attendants worked together to seat all the passengers and review with them the steps they should follow in case of an emergency. The pilots executed a series of complex, interdependent tasks that changed from the routine to the exceptional after the bird strike. As the emergency unfolded, air traffic controllers and airport personnel working the incident communicated with each other, identifying and readying alternative landing locations. When the plane ditched in the river, other groups swung into action: Firefighters, the Coast Guard, ferry boat crews, and search and rescue teams converged on the scene. The passengers, too, worked together as they escaped the doomed aircraft and waited in the freezing January weather for help to arrive.
This chapter examines these kinds of groups—ones whose members are working together to produce something or reach a goal—by asking three important questions. First, are we better together? Much of the work we do does not require collaboration, yet we often prefer to do that work with others rather than alone. Does the presence of other people facilitate our performance, and if it does, why? Second, do many hands make light the work? Certainly groups can do more than single individuals—what with more people to bear the load—but do we give our all when working in groups? Do groups bring out the best in us, or are our groups places where we can, and do, loaf? Third, when are two heads better than one? Do groups always outperform individuals, or does their superiority depend on the type of task the group is attempting? For example, are groups particularly good at finding creative, unusual solutions to problems, or do they tend to remain entrenched in their traditional ways of thinking? Chapters 11 and 12 examine groups working as teams and making decisions, respectively, but this chapter explores productivity and performance in all types of groups—from the simplest situation (two people working side by side on separate tasks) to more complex forms of productive interdependency.
10-1. Social Facilitation
When the last passenger on Flight 1549 was seated and the flight attendants closed the door, first officer Skiles completed the preflight checklist and taxied to the departure runway. The procedures Skiles followed were complex ones, but they were very familiar to him, for during his long career as a pilot he had flown thousands of take-offs at airports all over the United States. Skiles, however, was not alone in the cockpit. Captain Sullenberger was seated beside him, completing a second set of procedures required at take-off, but also monitoring Skiles’ actions. Did Sully’s presence influence Skiles’ performance?
10-1a. Performance in the Presence of Others
Norman Triplett’s 1898 study of people’s reactions to other people was one of the first experiments ever conducted in psychology. Triplett’s (1898) inspiration: bicycle races. In some events, cyclists raced alone and their performance was timed. Other events were competions with cyclists racing each other. In a third type of race, a rider was paced by a motor-driven cycle. Invariably, riders achieved their best times when they competed or they were paced, and they were slowest when racing alone. Triplett wondered why.
Triplett knew about drafting: riders are faster when biking with others because the lead cyclist creates a partial vacuum that pulls followers along while also breaking down wind resistance. But Triplett was also interested in identifying what he called “dynamogenic factors”:
The bodily presence of another rider is a stimulus to the racer in arousing the competive instinct; that another can thus be the means of releasing or freeing nervous energy for him that he cannot of himself release; and, further, that the sight of movement in that other by perhaps suggesting a higher rate of speed, is also an inspiration to greater effort. (p. 516)
To eliminate the possibility of drafting, he arranged for 40 children to perform a simple reel-turning task in pairs and when alone (see Figure 10.1). His study was a success, for he was able to confirm the positive impact of working in the presence of other people: social facilitation. The children in pairs outperformed those who worked alone (see Strube, 2005, for a reanalysis of Triplett’s data).
Figure 10.1
Details
The “competition machine” used by Triplett in his research examining the positive influence of competition on performance. Triplett is still recognized for his noteworthy contribution to the scientific study of groups, even though a reanalysis of his findings using modern statistics indicated the differences between the conditions he studied were not very substantial. In all likelihood, had he performed his study today instead of in 1898, his fellow researchers would have sent him back to his laboratory to find more convincing evidence of those mysterious dynamogenic factors (Strube, 2005).
Coaction, Audiences, and Inconsistencies
Triplett studied a coaction situation: people working in the presence of other people, but not necessarily interacting with one another. People digging separate holes in a field, taking a test in a classroom, or riding bicycles with friends are common coaction situations. But researchers soon discovered that social facilitation also occurs when individuals perform in the presence of an audience. One investigator, for example, studied the effects of an audience by recording how much weight men could lift when exercising. Men who were watched when they were working out could lift heavier weights (Meumann, 1904).
Other studies, however, did not confirm the “presence of people improves performance” effect. Social psychologist Floyd Allport (1920), for example, arranged for participants to complete tasks twice—once while alone in a small testing cubicle, and once with others at a table. To reduce competition, Allport cautioned participants not to compare their scores with one another, and he also told them that he himself would not be making comparisons. He found that people in groups produced more than isolated individuals, but their products were often lower in quality. Likewise, other researchers sometimes reported gains in performance through coaction or when an audience was watching, but they also documented performance decrements (Aiello & Douthitt, 2001).
Zajonc’s Resolution
Confusion reigned until Robert Zajonc (1965) explained why different studies yielded such divergent results. Some behaviors, he noted, are easier to learn and perform than others. These dominant responses are located at the top of the organism’s response hierarchy, so they dominate all other potential responses. Behaviors that are less likely to be performed are nondominant responses. Zajonc noticed that researchers documenting social facilitation studied well-learned or instinctual responses, such as lifting weights, bicycling, or eating rapidly. Studies involving novel, complicated, or unpracticed actions, such as solving difficult math problems or writing poetry, usually found little evidence of social facilitation.
Zajonc’s insight was that the presence of others increases the tendency to perform dominant responses and decreases the tendency to perform nondominant responses. If the dominant response is the correct or most appropriate response in a particular situation, then social facilitation occurs; people will perform better when others are present than when they are alone. If the task calls for nondominant responses, then the presence of other people interferes with performance (see Figure 10.2). Imagine that you must memorize some pairs of words. If the pairs are common associations, such as blue–sky or clean–dirty, then the task is an easy one, for which the dominant response is correct. Hence, your performance will be better if other people are present. If, however, you are trying to learn some uncommon associations—such as blue–dynamogenic or clean–nondominant—then you are required to make a nondominant response and an audience will hurt more than help.
Speed, Quantity, and Quality
Many studies, including those listed in Table 10.1, have confirmed the relationship between task complexity and social facilitation. Novice drivers perform more poorly when an audience is present in the car. Individuals dress more quickly when another person is present, provided they are putting on familiar articles of clothing. Right-handed people can write faster with that hand when another person is present, but even more slowly if they are trying to write with their left hand. Social facilitation has even been documented in other species. Cockroaches, horses, puppies, chickens, mice, rats, monkeys, armadillos, ants, beetles, and opossums are on the list of animals that show signs of increased performance in the presence of other members of their species (Clayton, 1978). When researchers reviewed hundreds of studies of over 24,000 humans meta-analytically, they concluded that social facilitation is most likely to occur on tasks where speed and quantity matter more than accuracy. So long as the task is a simple one, people tend to work more quickly when others are present, and the result is a small, but consistent, uptick in productivity. The presence of other people interferes with speed, however, when the task is complex, so other people significantly inhibited both the quantity and quality of their performance. Overall, the gains that occurred when people worked together on simple tasks were not as great as the losses that occurred when people worked on complex tasks (Bond & Titus, 1983).
Table 10.1 A Sampling of Empirical Demonstrations of Social Facilitation
Situation Findings
Dressing People performed a familiar task (taking off their own shoes and socks) faster if another person was in the room, and even faster if watched by an observer. They took longer to finish a less familiar task (putting on a robe that tied in the back) when another person watched them (Markus, 1978).
Driving Individuals seeking their license to drive an automobile took their driving test with only the tester or with another test-taker seated in the rear seat. Forty-nine percent of the applicants passed the test when alone, but only 34% passed when an audience was present (Rosenbloom et al., 2007).
Handwriting College students copied a word list using their dominant hand (easy task) or nondominant hand (hard task). The computer screen where they worked displayed an image of a favorite television personality or another character from the same program. If the task was easy, they wrote more words in the presence of their favorite character; if the task was difficult, the favorite character inhibited their performance (Gardner & Knowles, 2008).
Jogging The path taken by solitary women joggers passed by a woman who either watched them as they ran or sat facing away from them. Joggers accelerated when they encountered the watchful observer (Worringham & Messick, 1983).
Playing pool People playing pool were surreptitiously watched to identify skilled and unskilled players. When the observer moved near the pool table and openly watched their play, skilled players’ performance improved 14% but unskilled players’ performance deteriorated by more than 30% (Michaels et al., 1982).
Resource acquisition Children used a candy and fruit dispenser when alone, when observed by another child, or in the presence of another child also using a dispenser (coaction). Children worked harder (acquired more candy) in the coaction and observer conditions. Similarly trained chimpanzees worked harder only in the coaction condition (Engelmann, Herrmann, & Tomasello, 2016).
Security screening Individuals using a baggage security scanner screened luggage containing fewer items more quickly than luggage with multiple items, and this tendency was more pronounced when they were observed by another person. Presence of an observer did not influence the accuracy of their screening (Yu & Wu, 2015).
Speaking On a writing task most people (93%) produced more words when another person was present than when they were alone (Allport, 1920). When this study was replicated with individuals who stuttered when they spoke, 80% of the subjects produced more words when alone rather than with another person (Travis, 1928).
Is Social Facilitation a Uniquely Human Phenomenon?
Social facilitation is not limited to Homo sapiens: horses, puppies, chickens, mice, rats, monkeys, armadillos, ants, beetles, and opossums are on the list of species that show signs of increased performance in the presence of other members of their species. Even the lowly cockroach will work harder when surrounded by other cockroaches. As anyone who has surprised a roach in the kitchen late at night knows, cockroaches run from bright lights. So Zajonc and his colleagues (1969) designed two mazes with a start box near a light and a goal box hidden from the light. One maze was easy, even for a roach—just a straight runway from start to the goal. The second maze was more complex: the roaches had to turn to the right to reach their goal. Zajonc then timed how quickly 72 roaches (Blatta orientalis) completed the mazes when alone, when with another roach, or when watched by other cockroaches—although we cannot be sure the spectator roaches actually watched. They were sealed in small plastic boxes adjacent to the mazes, with air holes allowing air to circulate between spectator and subject.
Zajonc’s findings were consistent with the findings from studies of humans. In the simple maze, single roaches reached home base in an average of 40.6 seconds. Coacting roaches trimmed 7.6 seconds off this time, returning in just 33 seconds flat. This tendency reversed when the maze was complex: Single roaches crawled to the finish line 19.6 seconds faster than did coacting roaches. Roaches watched by an audience were the slowest contestants of all, but they were particularly slow when the maze was complex—taking nearly 2 minutes longer than single roaches.
10-1b. Why Does Social Facilitation Occur?
The situations studied by Triplett and Zajonc barely qualify as groups, for they involved strangers working on individualized tasks without any interaction, influence, shared identity, or common goals. Yet, even these circumstances were sufficient to trigger psychological and interpersonal processes that sometimes facilitated, and sometimes interfered with, performance (see Aiello & Douthitt, 2001; Monfardini et al., 2016).
Drive Processes
Zajonc coined the word compresence to describe the state of responding in the presence of others. Compresence, he hypothesized, touches off a basic arousal response in most social species “simply because one never knows, so to speak, what sorts of responses—perhaps even novel and unique—may be required in the next few seconds” when others are nearby (Zajonc, 1980, p. 50). Zajonc believed that compresence in and of itself elevated drive levels that triggered social facilitation when tasks were so easy that only dominant responses would be needed to perform them.
Zajonc’s drive theory uniquely predicts that social facilitation will occur even when all forms of social interaction, communication, and evaluation between the individual and the observer are blocked. Investigators tested this hypothesis by asking people to work on simple or complex tasks in the presence of an “observer” who was blindfolded and wore earplugs. Even though the observer could not interact with participants in any way, his mere presence still enhanced their performance when they worked on simple tasks and slowed their performance on complex ones (Schmitt et al., 1986).
Physiological Processes
Zajonc’s drive theory suggests that people react, physiologically, to the presence of people—but the magnitude of this change depends on the type of situation and on who is watching. For example, James Blascovich and his colleagues (1999), in their studies of the threat/challenge model, have verified that an audience triggers increases in cardiac and vascular reactivity. Blascovich’s team also discovered, however, that this arousal is physiologically very different when people work on an easy task rather than on a hard one. As noted briefly in Chapter 2, when we are confident we can master a problem or difficulty, we tend to display a challenge response. At the physiological level, we appear to be ready to respond to the challenge that we are facing (elevated heart rate and sympathetic nervous system activation). More daunting tasks, in contrast, may trigger a threat response; we display a stress reaction instead of a challenge response. The challenge response facilitates our performance, whereas the threat response inhibits it (Blascovich et al., 1999).
Neurological Processes
Social neuroscientists also trace social facilitation back to a physiological process, but locate that process in the brain itself. Since humans are a social species, we are neurologically prepared to monitor and respond to other individuals and groups. This so-called “social brain” includes mechanisms that sustain and promote our capacity to perceive and understand other people, and respond to them appropriately (e.g., Tremblay, Sharika, & Platt, 2017). When a task is a simple one, this activation of the attentional and reward centers of the social brain facilitate performance, but when the task is more difficult, the social brain interferes with performance (Monfardini et al., 2016). In one study that tested this hypothesis, researchers confirmed that the presence of another person activated a specific area of the brain associated with social cognition (the dorsolateral prefrontal cortex), and that this increase was associated with inhibited performance on complex, but not simple, tasks. This study found relatively few effects of other physiological processes, such as heart rate and blood pressure, on performance, prompting the investigators to conclude that it is the brain, and not the heart, that overreacts to the presence of others (Ito et al., 2011).
Motivational Processes
The 49-year-old Skiles earned his pilot license when he was 16. With over 20,000 hours logged as a commercial pilot, he was nearly as experienced as the plane’s captain, Sully. But Skiles had only recently completed his training for this particular aircraft, and this was his first flight without an instructor. Skiles and Sully did not know each other particularly well, for they had only begun working together as a flight crew four days earlier on the first leg of their seven-leg flight rotation. As Skiles maneuvered the plane along the tarmac, he knew that Sully was watching his every move.
Psychologist Nickolas Cottrell (1972) suggested that this evaluative pressure is one of the reasons why people tend to be more productive in the presence of others. His evaluation apprehension theory assumes that most of us have learned through experience that other people are the source of the rewards and punishments we receive. So we associate social situations with evaluation, and feel apprehensive whenever other people are nearby. This evaluation apprehension enhances our performance on simple tasks, but it becomes debilitating when we attempt more difficult projects. Cottrell thus believed that apprehension, and not the arousal response identified by Zajonc, is the source of social facilitation effects.
Sociologist Erving Goffman’s (1959) analysis of self-presentational processes, noted in Chapter 6, also underscores the motivational impact of impression management pressures. Self-presentation theory assumes each of us actively controls others’ impressions of us by displaying social behaviors that establish and maintain a particular social image, or face. We do not want the others to think that we possess negative, shameful qualities and characteristics, so we strive to make a good impression. Performance situations create self-presentational challenges for us, however, particularly when we feel we might fail. To avoid that embarrassment, we redouble our efforts when self-presentational pressures are strong—as they were in the cockpit when Skiles was piloting and Sully was watching (Bond, Atoum, & VanLeeuwen, 1996).
Researchers have tested, and in many cases confirmed, the primary hypothesis that derives uniquely from such motivational models—that any stimulus increasing the organism’s apprehension over future rewards or punishments should increase drive levels. When people find themselves in evaluative situations, they tend to perform dominant rather than nondominant responses (Seta et al., 1989). When, for example, individuals who were watched by an observer were told that the observer was evaluating them, their performance improved, but only when they were working on a simple task (Bartis, Szymanski, & Harkins, 1988). When people who had already failed once tried the task a second time, they performed worse when others were present (Seta & Seta, 1995). Also, situational factors that decrease evaluation apprehension, such as allowing for private responses, nonevaluative audiences, and the absence of a definable task that can be evaluated, often eliminate social facilitation effects (Henchy & Glass, 1968).
The presence of other people—even friends who we can count on for social support—can increase physiological reactivity in some circumstances. When women performed a difficult math test with a friend who was merely present—the friend could touch the participant’s wrist but was preoccupied with another task and was wearing a headset that blocked all sound—the participant’s cardiovascular responses were lowered. But when their friends watched them as they worked on difficult math problems, most people showed signs of physiological arousal rather than relaxation (Kamarck, Manuck, & Jennings, 1990). In fact, people are more relaxed when they are with their pets rather than with other people. Pets are an ideal source of social support, for they provide reassurance through their presence but they do not (we assume) evaluate their owner’s performance (Allen et al., 1991; Allen, Blascovich, & Mendes, 2002).
Other findings, though, do not support this emphasis on evaluation. Even when the companion refrains from attending to the individual in any way, social facilitation still occurs (Berger, 1981; Platania & Moran, 2001). Also, social facilitation occurs in animals that likely lack the capacity to feel nervous or embarrassed—rats, armadillos, and roaches, for example. Moreover, activities that involve little threat of evaluation, such as eating, drinking, or getting dressed, still show social facilitation effects.
Attentional Processes
Zajonc stressed drive levels, Cottrell underscored the importance of evaluation, but several cognitive theories have suggested that the presence of others changes people’s capacity to process information adequately. When people work in the presence of other people, they must split their attention between the task they are completing and the other person (Guerin & Innes, 1982). The presence of an audience may also increase individuals’ self-awareness, and, as a result, they may focus their attention on themselves and fail to pay sufficient attention to the task (Mullen & Baumeister, 1987).
Distractions, however, do not inevitably undermine performance. Distraction–conflict theory suggests that distraction interferes with the attention given to the task, but that these distractions can be overcome with effort. Therefore, on simple tasks that require dominant responses, the interference effects are inconsequential compared with the improvement that results from concentrating on the task so performance is facilitated. On more complex tasks, the increase in drive is insufficient to offset the effects of distraction, and performance is therefore impaired (Baron, 1986; Bond et al., 1996; Sanders, Baron, & Moore, 1978).
Consider, for example, the effect of an audience on people performing the famous (in psychology, at least) Stroop task. In the Stroop task, participants are shown a color name (e.g., Red or Blue) printed in a primary color (such as red or blue) and are asked to name the color of the ink. For example, if the word Red is printed in blue ink, the participant should answer blue. When the ink and the color word match, people have no problems. But when the ink and the color word are incongruent, reaction time and errors increase. But these errors decrease when individuals complete the task with others. The presence of others may work by helping people narrow their focus of attention and by filtering out the distracting color name cue (Huguet et al., 1999). The effect may also be due to
(a)
the extra cognitive demands imposed on participants by the presence of the observer and the need to evaluate the task itself (Klauer, Herfordt, & Voss, 2008) or
(b)
the increased attentional focusing on the task that is triggered by a threat of self-evaluation (Normand, Bouquet, & Croizet, 2014).
Cognitive Processes
When Sully announced “my aircraft” and took control of the mortally wounded Airbus A320, his mind was busy calculating any number of possible courses of action, factoring in the craft’s altitude and speed, distance to available airports, the hazards of a waterlanding, the number of civilians on the ground who could be injured, and so on. But the stress of the situation undoubtedly influenced how he processed that information, helping him to narrow his focus to concentrate on his best options while ignoring untenable alternatives.
Social psychologist Stephen Harkins’ (2006) mere-effort model (or, more precisely, the Threat-Induced Potentiation of Prepotent Reponses model) suggests that the gains and losses in performance we exhibit when we work on simple and complex tasks are due, in part, to changes in the way we process information. Harkins agrees with other researchers who note that evaluation usually triggers increases in effort: when we think we are being evaluated, we work harder. This increased effort causes us to concentrate more on ideas and information that are readily accessible to us, and if this information is relevant to the task at hand, then facilitation will occur. However, if this information is not relevant, then thinking about this information will inhibit our performance (Seitchik, Brown, & Harkins, 2016).
To test this hypothesis, Harkins (2006) measured the performance of individuals who worked on simple and complex versions of the Remote Associates Test—the RAT. Each item on the RAT consists of three words, and the test-taker’s task is to provide a single word that the three stimulus words have in common. A simple RAT item would be birthday–playing–shark, since the correct answer (card) is a close associate of all three words in the list. A complex RAT item, such as elephant–lapse–vivid, is more difficult because the correct answer (memory) is a remote rather than close associate, and so the test-taker must discard the close associates and search for more distant ones to solve the problem.
Harkins discovered that evaluative pressures improve performance on simple RATs but slowed performance on complex RATs, as evaluation apprehension theory would suggest. But he also discovered that this effect occurs because the evaluative pressure increased the availability of close associates (the “pre-potent responses”), which were only correct when people were working on the simple RAT. People could solve the more complex RAT items, but they needed time to move beyond the close associates that came to mind so easily.
Personality Processes
Sully, in the months following the crash, became a public figure: He appeared on talk shows, testified before Congress, accepted awards and accolades from various groups, and even threw out the first pitch at a couple of baseball games. He took it all in stride, rising to each occasion as he did when he landed Flight 1549 on the Hudson.
Social orientation theory suggests that people differ in their overall orientation toward social situations, and these individual differences in social orientation predict who will show facilitation in the presence of others and who will show impairment. According to this theory, individuals who display a positive orientation are so self-confident that they react positively to the challenge the group may throw their way. Others, in contrast, display a negative orientation. They approach social situations apprehensively, for they feel inhibited and threatened by other people. People may be capable of adopting either orientation in a given situation, but people like Sully are naturally positive in their orientation toward tasks. Others, in contrast, possess personality traits that prompt them to be more negative, such as low self-esteem, self-consciousness, anxiety, and neuroticism. A meta-analysis of previous studies of social facilitation, focusing only on those studies that included measures that might be indicators of participants’ degree of positive or negative orientation, supported the theory. Individuals with qualities that suggested their social orientation was positive usually showed social facilitation effects, whereas those with a negative orientation showed a social interference effect (Uziel, 2007, 2010, 2015).
10-1c. Conclusions and Applications
Social facilitation occurs because humans, as social beings, respond in predictable ways when joined by other members of their species (see Table 10.2). Some of these reactions, as Zajonc suggested, are very basic ones, for the mere presence of other people elevates drive levels. But arousal becomes more substantial when group members realize that the people around them are evaluating them and might form a negative impression of them if they perform badly. Cognitive and personality mechanisms that govern how individuals process information and monitor the environment also come into play when people work in the presence of others. As the following examples illustrate, these physiological, motivational, cognitive, and personality processes influence group members’ reactions across a wide range of performance settings.
Table 10.2 Theoretical Explanations of Social Facilitation and Supporting Evidence
Processes Theoretical Assumptions Evidence
Drive processes The mere presence of others elevates drive levels, resulting in social facilitation when tasks are so easy that only dominant responses are needed to perform them (Zajonc, 1965).
Social facilitation occurs when other people are “merely” present.
Many species perform basic tasks more efficiently in the presence of other species’ members.
Physiological and neurological processes (e.g., threat/challenge theory) Humans exhibit increased physiological arousal and heightened social attentiveness in the presence of other humans, and these reactions sometimes facilitate performance (e.g., Blascovich et al., 1999).
People show signs of physiological arousal when others are present.
Brain monitoring studies confirm that the cortical regions responsible for processing social information are activated in the presence of other people.
Motivational processes (e.g., evaluation apprehension theory and self-presentation theory) Through experience, people learn to associate the presence of others with evaluation; this evaluation apprehension facilitates performance on well-learned tasks (Cottrell, 1972).
The presence of others is facilitative only when the observers can evaluate the quality of the performance.
Facilitative effects are strongest when individuals are striving to make a good impression.
Cognitive processes (e.g., distraction–conflict theory and mere-effort model) When others are present, attention is divided between the other people and the task; attentional conflict increases motivation, which facilitates performance so long as the task is a simple one (Baron, 1986; Harkins, 2006).
Recall is poorer when a stimulus is presented in the presence of others, suggesting others are distracting.
Facilitation is reduced if the others in the situation are not noticed.
The presence of others improves performance on interference tasks (e.g., the Stroop Task and the RAT ).
Personality processes (e.g., social orientation theory) Individuals who display a positive interpersonal orientation and are more motivated to present themselves positively to others are more likely to show social facilitation effects (Uziel, 2007).
The presence of others improves performance among individuals with high self-esteem and low neuroticism.
Those with an attention-seeking tendency (exhibitionism) perform better than self-conscious individuals in coaction settings.
Prejudice and Social Facilitation
Prejudices are deeply ingrained negative attitudes about the members of other groups. Such prejudices as racism and sexism are increasingly recognized as unfair and socially inappropriate, so individuals who are prejudiced often try to keep their prejudices to themselves to avoid being labeled a racist or sexist. But prejudice is often a well-learned, dominant response; ironically, the presence of other people may lead individuals to express even more biased opinions when they are in public rather than in private. The presence of others may work to facilitate prejudice, rather than keep it in check (Lambert et al., 1996, 2003).
Eating in Groups
The presence of other people facilitates one of the most dominant of basic responses in humans: eating. Most people report that they prefer to eat with others rather than dine alone (Higgs & Thomas, 2016). When researchers ask people to keep track of how much and with whom they eat, they usually find that people eat more—sometimes 40–50% more—when they dine in groups (Herman, 2015). As meals eaten by groups are longer in duration than those eaten by solo individuals, people have more opportunity to keep eating when in groups than alone. Watching someone else eat also increases social imitation of the eating response. When the participants in one study witnessed another person eating 20 soda crackers, they ate far more crackers themselves than did participants who saw someone eat only one (Nisbett & Storms, 1974). People even seem to prepare relatively larger portions for meals to be eaten in groups than individually, as if they anticipate that the group members will be able to consume more than they would if alone. So long as the group does not include a substantial portion of dieters, the group may continue to eat until all the available food is consumed. Solitary eaters are more likely to eat only until they are sated (Herman, Roth, & Polivy, 2003). Larger groups trigger greater increases in eating, although at a decreasing rate, similar to response patterns suggested by social impact theory (Latané, 1981).
Groups do not always facilitate eating, however. The social facilitation of eating is weak when co-eaters are strangers or disliked, and strongest when people dine with families and friends. The social facilitation of eating is also limited to coaction rather than audience situations. People eat more when others with them are eating, but they tend to eat less when the other people who are present are observing them (Herman, 2015).
Electronic Performance Monitoring
Social facilitation is not limited to face-to-face, or collocated, group settings. The presence of others in a virtual sense—made possible when people join with others via computers, telephones, or other communication systems—can also enhance performance on simple tasks but undermine performance on complicated ones. Social psychologist John Aiello, for example, drew on studies of social facilitation in his analyses of electronic performance monitoring, or EPM (Aiello & Douthitt, 2001; Dohthitt & Aiello, 2001). When workers use their computer to enter data, communicate with one another, or search databases for stored information, their activity can be monitored automatically; as many as 75% of all companies in the United States use that data to monitor the performance of their employees (Zweig & Scott, 2007).
Does EPM enhance performance, or does it create so much evaluation anxiety that performance suffers? Aiello found that EPM may enhance employees’ productivity, but in ways that are consistent with social facilitation effects. He studied people working on a data entry task. Some were alone, some were working with others, and some were members of a cohesive group. Aiello discovered that EPM enhanced the performance of highly skilled workers, but interfered with the performance of less skilled participants. Monitoring also increased workers’ feeling of stress, except among those who were part of a cohesive work group. Individuals responded more positively to monitoring when they believed that they could turn off the monitoring and that only their job-related activities were being monitored, as well as when they had the opportunity to participate in decisions about the use of the monitoring system (Zweig & Scott, 2007).
Social Facilitation in Educational Settings
Much can be learned when one is alone, but many learning activities are group activities. Students can read and study in isolation, but more frequently a class of learners is assembled in one place with the hope that learning will occur en masse.
But even though learning in a social context is a common practice, the presence of other people may actually inhibit the acquisition of new concepts and skills. Others can be distracting, and, during the early phases of learning, this distraction can interfere with overt and covert practicing. When the participants in one project needed to learn a list of words, they were too embarrassed to rehearse the material by saying it aloud, and their performance suffered (Berger et al., 1981, 1982). Studies of athletes acquiring new skills, students learning a second language, and clinicians developing their therapeutic skills have indicated that learning proceeds more rapidly, at least initially, when learners work alone (Ferris & Rowland, 1983; MacCracken & Stadulis, 1985; Schauer, Seymour, & Geen, 1985).
Zajonc (1965, p. 274), however, suggests that once students have learned their skills well, then they should perform with others present if possible (Utman, 1997). He recommends the student:
study all alone, preferably in an isolated cubicle, and arrange to take his examinations in the company of many other students, on stage, and in the presence of a large audience. The results of his examination would be beyond his wildest expectations, provided, of course, he had learned his material quite thoroughly.
Do You Prefer to Learn Alone or in a Group?
Many educators take advantage of the motivational and pedagogical benefits of groups in their teaching. When students work and learn together in groups, they can pool their knowledge and abilities, give each other feedback, and tackle learning tasks too overwhelming to face alone. Group assignments, even if they are not the most efficient way to learn specific facts and information, help students develop a rare interpersonal skill—that of working effectively with others in groups.
Group approaches to learning, however, are not enthusiastically embraced by everyone. One survey of college students found that many had taken classes that used group-based experiences, but only 32% of the students rated their small-group experiences positively in terms of learning outcomes (Hillyard, Gillespie, & Lettig, 2010). For every person who enjoys studying in groups is another person who has spent too much time in a distracted, unproductive group. Are you, personally, pro or con learning in groups?
Instructions: For each of the statements, indicate the degree to which you agree or disagree by circling a number where:
1 = Strongly disagree
2 = Moderately disagree
3 = Neither agree nor disagree
4 = Moderately agree
5 = Strongly agree
1. I prefer to study for classes by myself. 1 2 3 4 5
2. Study groups waste too much time discussing things unrelated to class. 1 2 3 4 5
3. I mainly depend on myself, rarely on others when it comes to my learning. 1 2 3 4 5
4. Most study groups fail due to a lack of leadership in the group. 1 2 3 4 5
5. I have a negative opinion of study groups. 1 2 3 4 5
Scoring: Add your answer to all five items to yield a total. A score of 15 is close to the average score, but if your total is less than 11 you are relatively positive about studying in groups; if your score is above 19: you are no fan of learning collectively. But there is hope. Now that you have far more insight into groups and their dynamics, you can take steps to maximize your next learning group’s efficiency.
10-2. Social Loafing
Modern air travel requires not only the services of the group of pilots who fly the planes, but dozens of other groups also who handle service, ticketing, security, maintenance, transportation, and so on. The work of these groups is often outstanding, but anyone who has traveled has likely encountered a group that was neither efficient nor effective. Captain Sullenberger, describing his own experience in the industry, writes: “The gate agent hasn’t pulled the jetway up to the plane in time. The skycap is supposed to bring the wheelchair and hasn’t…. The caterer hasn’t brought all the first-class meals…. You get tired of constantly trying to correct what you corrected yesterday” (Sullenberger, 2009, pp. 158–159). Some groups, it seems, are not so intent on maximizing their productivity.
10-2a. The Ringelmann Effect
Max Ringelmann (1913), a nineteenth-century French agricultural engineer, was one of the first researchers to study group productivity. Ringelmann’s questions were practical ones: How many oxen should be yoked in one team? Should you plow a field with two horses or three? Can five men turn a mill crank faster than four? But Ringelmann, instead of speculating about the answers to these questions, set up teams of varying sizes and measured their collective power.
Productivity Losses in Groups
Ringelmann’s most startling discovery was that workers—and that includes horses, oxen, and men—all become less productive in groups. A team pulling a rope in a tug-of-war contest is stronger than a single opponent but the group does not usually work at maximum efficiency. When Ringelmann had individuals and groups pull on a rope attached to a pressure gauge, groups performed below their predicted potential productivity. Working alone, a single person could pull an average of 85.3 kg. But, when working with six other people, they could pull 390 kg, which is only 65 kg per person: a reduction of nearly 24% of their pulling power. Larger groups were even less productive, with each member pulling only 61.4 kg per person (see Figure 10.3). Groups certainly outperformed individuals, but as more and more people were added, the group became increasingly inefficient. To honor its discoverer, this tendency for groups to become less productive as group size increases is now known as the Ringelmann effect (Ingham et al., 1974; Kravitz & Martin, 1986 present an excellent summary and interpretation of Ringelmann’s work.)
Ringelmann believed that this reduction in productivity had two sources. First, coordination losses, or “the lack of simultaneity of their efforts,” introduced inefficiencies into each group (Ringelmann, 1913, p. 9). Even on a simple task, such as rope pulling, people tend to pull and pause at different times, resulting in some process loss and a failure to reach their full productive potential. Ringelmann’s groups often sang together in an attempt to synchronize their efforts and minimize coordination losses. Second, motivation losses were also sapping group productivity: People did not work as hard when they were in groups rather than alone. After watching a group of prisoners turning the crank of a flour mill, for example, he noted that their performance was “mediocre because after only a little while, each man, trusting in his neighbor to furnish the desired effort, contented himself by merely following the movement of the crank, and sometimes even let himself be carried along by it” (p. 10; translation from Kravitz & Martin, 1986, p. 938). This reduction of effort by individuals working in groups is now known as social loafing (Williams, Harkins, & Latané, 1981).
Many Hands Make Light the Work
Bibb Latané, Kipling Williams, and Stephen Harkins disentangled the effects of both coordination loss and social loafing in a series of cleverly designed studies. They told the men they recruited for their groups that they were researching “the effects of sensory feedback on the production of sound in social groups” and that all they needed to do was to cheer as loudly as they could. They asked the participants to wear blindfolds and headsets, so their performance would not be influenced by “the effects of sensory feedback” (1979, p. 824). They then asked participants to shout as loudly as they could while the headsets played a stream of loud noise. Consistent with the Ringelmann effect, groups of participants made more noise than individuals, but groups failed to reach their potential. When the participants were tested alone, they averaged a rousing
9.22
dynes/cm
2
(about as loud as a pneumatic drill). In dyads, each participant shouted at only 66% of capacity and in six-person groups at 36% of capacity. This drop in productivity is charted in Figure 10.4 (Latané, Williams, & Harkins, 1979, Experiment 2, p. 826; see also Harkins, Latané, & Williams, 1980; Williams et al., 1981).
Social loafing and coordination losses in groups. Latané and his colleagues disentangled the two major causes of productivity losses in groups by leading people to think they were working in groups when they actually were not. The people in these “groups” (labeled “Pseudogroups productivity”) suffered from motivation loss, but not from coordination loss since they were actually working alone. The shaded portion represents motivation loss (social loafing) and the unshaded portion represents coordination loss.
They combine to create the Ringelmann effect.
But how much was this drop in productivity due to social loafing and how much due to coordination problems? Latané and his colleagues separated out these sources of process loss by testing noise production in “pseudogroups.” In these conditions, participants were led to believe that either one other participant or five other participants were shouting with them, but in actuality, they were working alone. (The blindfolds and headsets made this deception possible.) Thus, any loss of production obtained in these pseudogroup conditions could not be due to coordination problems, because there were no other group members shouting. Instead, any decline in production could only be blamed on the reduced effort brought about by social loafing. As Figure 10.4 indicates, when participants thought that one other person was working with them, they shouted only 82% as intensely. If they thought that five other persons were shouting, they reached only 74% of their capacity. These findings suggest that even if work groups are so well organized that virtually all losses due to faulty coordination are eliminated, actual productivity will not equal the potential productivity due to social loafing.
10-2b. Causes of and Cures for Social Loafing
People carrying out all sorts of physical and mental tasks—including brainstorming, evaluating employees, monitoring equipment, interpreting instructions, and formulating causal judgments—have been shown to exert less effort when they combine their efforts in a group situation. Even worse, loafing seems to go unrecognized by group members. When people in groups are asked if they are working as hard as they can, they generally claim that they are doing their best, even though the objective evidence indicates that they are loafing. Evidently, people are not aware that they are loafing, or they are simply unwilling to admit it (Karau & Williams, 1993). Fortunately, researchers have identified a number of steps that can be taken to reduce the level of social loafing in a group.
Increase Identifiability
Studies of social loafing suggest that people are less productive when they work with others. But studies of social facilitation, discussed earlier in this chapter, find that people are more productive when others are present (at least when the task is easy). Which is it?
Both. When people feel as though their level of effort cannot be ascertained because the task is a collective one, then social loafing becomes likely. But when people feel that they are being evaluated, they tend to exert more effort and their productivity increases. If the task is an individualistic one, and is easy, social facilitation occurs. But when group members are anonymous and their contributions are unidentifiable, the presence of others reduces evaluation apprehension, and social loafing becomes more likely (Harkins & Szymanski, 1987, 1988; Jackson & Latané, 1981).
Researchers illustrated the importance of evaluation by asking the members of four-person groups to generate as many ideas as possible for a common object. The participants did not discuss their ideas out loud but simply wrote them on slips of paper. Some of the participants thought that their ideas were individually identifiable, whereas others thought that their ideas were being collected in a common pool. Moreover, some participants believed that everyone was devising uses for the same object, but others thought that each group member was working with a different object. In this study, loafing occurred not only when ideas were pooled, but also when the participants believed that their individual outputs were not comparable or could not be evaluated (Harkins & Jackson, 1985). Loafing was reduced when each individual member’s output was identifiable (Hardy & Latané, 1986; Kerr & Bruun, 1981; Williams et al., 1981).
How Many Pickles Could a Pickle Packer Pack If Identifiable?
Many groups work on tasks as a team, and so member’s individual inputs are often inseparable: who did what is not clearly known. This intermixing can result in a loss of productivity, as Kipling Williams and his colleagues (1981, p. 310) explain in their analysis of performance procedures in the pickle processing plant:
[In a pickle packing plant] a key job is stuffing dill pickle halves into jars. Only dill halves of a certain length can be used. Those that are too long will not fit and those that are too short will float and dance inside and look cheap and crummy. The dill halves and the jars are carried on separate high-speed conveyor belts past the contingent of pickle stuffers. If the stuffers don’t stuff quickly enough, the jars pile up at the workers’ stations while they look for pickles of the appropriate length, so stuffers have a great temptation to stuff whatever pickles come readily at hand. The individual outputs of the stuffers are unidentifiable, since all jars go into a common hopper before they reach the quality control section. Responsibility for the output cannot be focused on any one worker. This combination of factors leads to poor performance and improper packing. The present research suggests making individual production identifiable and raises the question, “how many pickles could a pickle packer pack if pickle packers were only paid for properly packed pickles?”
Minimize Free Riding
Thousands of people listen to public radio without making a contribution when the radio asks for donations. Some audience members do not clap during the call for an encore because they know their applause will not be missed. Many students avoid group projects where the entire group receives the same grade, because inevitably one or more members of the group will not do their share of the work (Hoffman & Rogelberg, 2001).
All these situations invite free riding—members doing less than their share of the work because others will make up for their slack. Although norms of fairness warn members to do their part, if they feel that the group does not need them or their contribution, they will be tempted to free-ride. They are also more likely to free ride if they suspect the other group members aren’t working very hard. Rather than looking like a “sucker” by working harder than the others, group members reduce their efforts to match the level they think other group members are expending. This sucker effect is strongest when they feel that their fellow group members are competent but lazy (Hart, Bridgett, & Karau, 2001).
The effects of free riding can be minimized by reducing the size of the group, strengthening the group’s performance norms, sanctioning those who contribute too little, and increasing members’ sense of indispensability: individuals who feel that their contribution to the group is unique or essential for the group’s success work harder (Kerr & Bruun, 1983). For example, when researchers studied the performances of swimmers competing on relay teams in the 2008 Olympics, they discovered that athletes who held the anchor spot on the team showed performance gains when swimming in the relay. Swimmers who started the relay, swimming in either the first or second positions, did not swim any better than they did when competing as individuals (Hüffmeier & Hertel, 2011).
Set Goals
Groups that set clear, challenging goals outperform groups whose members have lost sight of their objectives (Kleingeld, van Mierlo, & Arends, 2011). When truck drivers who hauled logs from the woods to the mill were initially told to do their best when loading the logs, the men only carried about 60% of what they could legally haul (Latham & Baldes, 1975). When the drivers were later encouraged to reach a goal of 94% of the legal limit, they increased their efficiency and met this specific goal. In a study of groups generating ideas, members were more productive when they had a clear standard by which to evaluate the quality of their own work and the group’s work (Harkins & Szymanski, 1989). Members of sports teams with norms that stress goal pursuit (e.g., “we don’t give up during adversity in a competition”) loaf less than teams with lower performance standards (Høigaard, Säfvenbom, & Tønnessen, 2006, p. 222). The group’s goals should also be challenging rather than too easily attained. The advantages of working in a group are lost if the task is so easy that it can be accomplished even if the group loafs, so care should be taken to set the standards high—but not so high that they are unattainable (Hinsz, 1995; Weldon & Weingart, 1993).
Increase Involvement
Sullenberger (2009) was very involved in his work as a pilot. Hardworking, attentive to detail, and serious about providing quality service, his wife described him as a “pilot’s pilot” (p. 276). People like Sully do not loaf in groups. When researchers first screened people on a set of questions that measured their approach to work—the Protestant Ethic Scale—they discovered people with high scores loafed very little. Such sentiments as “People who fail at a job have usually not tried hard enough” and “There are few satisfactions equal to the realization that one has done his best at a job” are antithetical to not doing one’s share of the work (Smrt & Karau, 2011). Individuals who enjoy competition and working with others in groups are also less likely to loaf (Stark, Shaw, & Duffy, 2007).
In general, the more engaged people are in the group or the group’s work, the less likely they will loaf. So long as the competition remains “friendly,” group members may persevere with much greater intensity when they are vying with others in the group for the best score (Hinsz, 2005). Challenging, difficult tasks reduce loafing, but so do those that will determine group members’ personal outcomes—either by reward or by punishment (Brickner, Harkins, & Ostrom, 1986; Shepperd, 1993, 1995). Social loafing is also reduced when rewards for successful performance are group-based rather than individually based—so long as the group is not too large in size (DeMatteo, Eby, & Sundstrom, 1998) and the reward is divided nearly equally among all the group members (Honeywell-Johnson & Dickinson, 1999; Liden et al., 2004). When large groups are split into smaller ones, members loaf less (Pentland, 2014).
Involvement may even prompt group members to compensate for the expected failures or incompetencies of their fellow group members by expending extra effort. Williams and Karau (1991) documented social compensation by convincing individuals that their group’s task was a meaningful one, but that the motivation of other group members was in doubt (apparently because one of the other experimenters considered the research topic to be boring). Participants were also led to expect that their partners were either skilled or unskilled at the task. Williams and Karau discovered that people worked hardest when the task was meaningful and the members believed that their coworkers’ ability was minimal. A field study of loafing in a classroom setting even suggests that a high level of involvement may trump the sucker effect. If students’ grades were on the line, when they discovered that one of their group members was a loafer, they tended to work harder themselves, rather than reducing their own effort to look less like a sucker (Liden et al., 2004).
Increase Identification with the Group
Social identity theory also suggests a way to reduce loafing: increase the extent to which group members identify with their group or organization (Haslam, 2004). Social identity theory suggests that the difference between a hardworking group and a loafing group is the match between the group’s tasks and its members’ self-definitions. If people are working together but the group and its tasks have no meaning to them, they care very little if their group succeeds or fails. But when individuals derive their sense of self and identity from their membership in the group, then social loafing is replaced by social laboring as members expend extra effort for their group. Individuals sometimes work hard when they think “This task is important to me,” but they are likely to work even harder when they think “This task is important to us” (Haslam, 2004).
Increasing identification with the group eliminates social loafing even in relatively short-term groups. When researchers assembled volunteers into three-person groups and had them complete two time trials—one short and one long—on stationary bicycles, they cycled more slowly in groups than they did when they were tested individually. But when the groups were led through a series of identity-building activities that included generating a team name and slogan before the trials, results were more impressive. The groups cycled just as quickly in their group as they had when cycling individually—so no social loafing—on the shorter trial, and they even outperformed their individual scores on the longer trials—exhibiting social laboring (Høigaard et al., 2013).
10-2c. The Collective Effort Model
Karau and Williams’s (1993, 2001) collective effort model (CEM) provides a comprehensive theoretical framework for understanding the causes and cures of social loafing. Drawing on classic expectancy-value theories of motivation, they suggested that two factors determine group members’ level of motivation: their expectations about reaching a goal and the value of that goal. Motivation is greatest when people think that the goal is within their reach (expectations are high) and they consider the goal to be valuable. Motivation diminishes if expectations are low or individuals do not value the goal. Working in a group, unfortunately, can diminish both expectations about reaching a goal and the value that is placed on that goal. In groups, the link between our effort and the chance of success is ambiguous. Even if we work hard, others may not, and the group may fail. Moreover, even if the group does succeed, we personally may not benefit much from the group’s good performance. Earning a good grade on a project completed by a group may not be as satisfying as earning a good grade on a project that we complete working on our own.
Karau and Williams tested the CEM’s predictions in a meta-analysis. Their review of 78 studies supported their basic theoretical contention that loafing is reduced if individuals’ expectations for success are high and they feel that the goal they are seeking is a valuable one. They also identified a number of other consistencies that emerged across studies. For example, loafing was greater among men than women, in Western countries compared to Eastern countries, and for simple tasks rather than complex tasks.
Is Your Group at Risk for Social Loafing?
Even in the best groups members may not be performing up to their potential; not because members are having difficulty working as a team, but because social loafing steals away their motivation. Asking members if they are working as hard as they can will not provide you with valid information about loafing, since most people don’t notice that they are not working as hard in the group as they would if they were alone. Instead, look for the warning signs of lost motivation listed here.
Instructions: From the many groups to which you belong, identify the one where members are working to achieve a collective goal of some type, such as a team at work, a study group, or even a group of friends planning social event. Then put a check by each item that accurately describes your experiences in this group.
Identifiability
People who work hard in this group don’t get credit for that.
This group doesn’t keep track of who does what.
Most people in this group don’t have any specific responsibilities.
Free-riding
Some members are not contributing very much to this group.
We have a couple of members who do tons of work for the group.
Even if some of us don’t work hard, our group will still reach its goals.
Goals
This group’s goals are not clearly defined.
It’s difficult to tell when we are making progress toward our goals.
This group’s goals are not challenging at all.
Involvement
This group isn’t one that means very much to its members.
It’s hard to predict who will show up for this group’s meetings and events.
If a group member falls behind in this group, that’s his/her problem.
Identification
Hardly anyone in the group feels strongly about the group’s purpose.
The members don’t identify strongly with this group and its goals.
No one really cares if this group is a great success.
Scoring: If you checked two of the three indicators in any one set, your group is displaying signs that its productivity may be limited by social loafing. What’s the solution? Use the methods that researchers have identified as effective (e.g., increased identifiability, involvement, indepensibility, and identification) to counteract these performance-limiting conditions, focusing on the problem areas you identified on this checklist.
10-3. Working in Groups
Groups tend to lose some of their productivity due to social loafing, but they usually outperform individuals. A lone individual in a tug-of-war with a group will lose. Individuals racing each other will run faster than they would if racing against the clock. A group taking a multiple-choice test will probably get a higher score than an individual taking the same test. Three flight attendants will be able to seat and service 150 passengers far more effectively than only one could.
But how well do groups perform on more complex tasks that require coordination and collaboration? Companies and businesses must monitor, regulate, and organize the activities of hundreds of employees—should they organize their workers into teams? When quality matters, will a single, dedicated craftsperson build a more beautiful product than a work crew that must plan each action as raw construction materials are transformed into a finished product? Mountaineers can climb alone, but can they reach the highest peaks only by working with others? When do groups outperform individuals?
10-3a. The Process Model of Group Performance
Ivan Steiner (1972), in his classic work Group Process and Productivity, drew on the concept of process loss to predict when groups will perform well or poorly. Steiner recognized that groups have great potential, for their resources outstrip those of any single individual. But Steiner also realized that groups rarely reach their full potential because no group can perfectly coordinate its resources, members, and processes. When individuals work by themselves, their performance depends strictly on their personal resources, including their talents, skills, and effort. But when individuals join together to work in groups, their performance depends on each individual’s resources plus the interpersonal processes that determine how these resources are combined. Even Sully and Skiles—experienced, well trained, and highly motivated—did not coordinate their actions perfectly during the emergency. In consequence, their potential productivity did not match their actual productivity.
Potential Productivity ≠ Actual Productivity
In Steiner’s (1972, p. 8) model, task-related processes are the steps group members take as they complete a task, including “all those intrapersonal and interpersonal actions by which people transform their resources into a product, and all those nonproductive actions that are prompted by frustration, competing motivations, or inadequate understanding.” Although a group’s potential productivity (PP) can be predicted by determining if its resources match the requirements of the task it is attempting, process loss (PL)—the grit in the interpersonal machinery of a working group—determines how effectively the group makes use of its resources. Steiner’s “law” of group productivity
AP
=
PP
-
PL
predicts that actual productivity (AP) is determined by a group’s potential productivity (PP) less all the process losses (PL) the group experiences. Some groups may have great potential: Members are highly motivated, well trained, and they have all the skills the group will need to master the task. Yet, even though the group has the potential to be successful, it may still fail because it fails to combine these resources effectively or efficiently.
Task Demands
Steiner recognized that many factors combine to determine a group’s potential productivity, but he highlighted one factor over all others: the type of task the group is attempting. A group working on an assembly line, for example, must combine members’ efforts in ways that differ from the combination process used by a team playing baseball or pilots flying a commercial jetliner. Some tasks, Steiner explained, require high levels of coordinated activity but others do not; even if group members make little or no attempt to adapt their actions to match those of others, the group will still succeed. Steiner called the combination processes dictated by the problem or group activity the task demands and suggested that they vary depending on the divisibility of the task, the type of output desired, and the combination rules required to complete the task (see Table 10.3 ).
Table 10.3 A Summary of Steiner’s Taxonomy of Tasks: Types, Qualities, and Examples
Divisibility: Can the task be broken down into subtasks?
Divisible The task has subcomponents that can be identified and assigned to specific members.
Playing a football game
Preparing a six-course meal
Unitary The task does not have subcomponents.
Pulling on a rope
Reading a book
Quantity versus quality: Is quantity produced more important than the quality of the performance?
Maximizing Quantity: The more produced the better the performance.
Generating many ideas
Lifting a great weight
Scoring the most goals
Optimizing Quality: A correct or optimal solution is needed.
Developing the best answer
Solving a math problem
Interdependence: How are individual inputs combined to yield a group product?
Additive Individual inputs are added together.
Pulling a rope
Shoveling snow
Compensatory A decision is made by averaging together individual decisions.
Estimating an ox’s weight by asking three people to guess and averaging their averaging the guesses
Averaging ratings of job applicants
Disjunctive The group selects one solution or product from a pool of members’ solutions or products.
Picking one answer to a math problem to be the group’s answer
Letting one art project represent the entire school
Conjunctive All group members must contribute to the product for it to be completed.
Climbing a mountain
Eating a meal as a group
Discretionary The group decides how individual inputs relate to the group product.
Deciding to shovel snow together
Choosing to vote on the best answer to a problem
SOURCE: Adapted from Group Processes and Productivity by I. D. Steiner. © 1972 by Academic Press.
Divisibility: Some tasks are divisible—they can be broken down into subtasks that can be assigned to different members—whereas other tasks are unitary. Building a house, planting a large garden, or working a series of math problems by assigning one to each group member are all divisible tasks, because the entire task can be split into parts. unitary tasks, however, cannot be divided: Only one painter is needed for a small closet in a house, only one gardener can plant a single seed, and only one person is needed to solve a simple math problem.
Quantity vs. quality: Some tasks call for a high rate of production (maximization), whereas others require a high-quality, correct outcome (optimization). With maximizing tasks, quantity is what counts. In a relay race, tug-of-war, or block-stacking problem, performance depends on sheer quantity; the emphasis is on maximal production. For optimizing tasks, a good performance is the one that most closely matches a predetermined criterion. Examples of optimizing tasks include estimating the number of beans in a jar or coming up with the best solution to a problem.
Interdependence: Members’ contributions to the group task can be combined in different ways. On an assembly line, for example, the members perform a specific task repeatedly, and the product is finished when each member has made his or her contribution. The members of a rock band, in contrast, all play and sing together, so each member’s contribution must mesh with the other members’ contributions. Steiner (1972) describes five basic combinatorial strategies: additive, compensatory, disjunctive, conjunctive, and discretionary.
By taking into account the type of task the group is attempting, the performance of groups relative to individuals can be predicted with more accuracy.
10-3b. Additive Tasks
On the surface, additive tasks are the easiest types of tasks for a group to complete. Since they are both divisible and maximizing, group members need only add their contribution to the group’s output, so coordination demands are minimal. So long as each group member can perform his or her assignment—such as pulling on a rope, editing an online encyclopedia, cheering at a football game, developing two slides for a PowerPoint presentation, clapping after a concert, responding to customer complaints in a call center, or raking leaves in a yard—the productivity of the group will probably exceed the productivity of a single individual.
Studies of both social facilitation and social loafing, however, warn that working on additive tasks is more complicated than it seems. If working in the presence of others, people may perform their piece of the additive task particularly well—but only if their subtask is a relatively simple one. People shucking green beans together on the front porch may work more effectively than individuals working separately, but once the task gets more challenging, the benefits of social facilitation will likely be negligible. Social loafing is also likely given the structure of an additive task. Adding more and more members will increase a group’s productivity when it works on an additive task, but at an ever decreasing rate of gain.
10-3c. Compensatory Tasks
When groups attempt compensatory tasks, the members average their individual judgments or solutions together to generate an answer. For example, each one of the passengers of Flight 1549 could be contacted and asked to estimate how long it took to rescue them from the cold waters of the Hudson after the crash. The estimates could then be averaged to generate a group judgment, which could be compared to the actual time taken: only 24 minutes (Miracle on the Hudson Survivors, 2009).
The Wisdom of Crowds Effect
Legendary nineteenth-century polymath Francis Galton was surprised by the accuracy of groups when making compensatory decisions. Known for his studies of intelligence, Galton questioned whether a group could possibly make more accurate judgments than an expert. He had the opportunity to test his hypothesis when he came across a “Guess the Weight of an Ox” contest at a local fair. Each contestant estimated the ox’s weight, and the person who came closest to the ox’s actual weight won a prize. Galton took the estimates home and examined them, expecting that the crowd would be far off the mark. Yet, the weight of the ox was 1,198 pounds, and the average of the judgments of the 800 contestants was 1,197, confirming the wisdom of the crowd effect (Surowiecki, 2004). Some people overestimated the ox’s weight, but others underestimated, so the group judgment, which was an average of all the estimates offered, was more accurate than the judgments made by experts and by most of the individuals.
Crowds are wiser than individuals for at least two reasons. First, the compensatory method is relatively immune to group process loss caused by poor coordination, loafing, or undue influence of the persuasive but unwise. In face-to-face groups, those who are well respected by the group—but not necessarily any better informed—often sway the group’s decision. They do not when groups work on compensatory tasks—provided the group members make their judgments independently of others. Second, a statistically derived group score is more accurate because it is based on multiple measures. When single individuals make multiple estimates, and their estimates are averaged, their judgments are also more accurate (Brennan & Enns, 2015). Because of the importance of basing the final estimate on a sufficient number of responses, the compensatory method requires a large enough number of judgments to compensate for any extreme judgments.
Swarm Size and Problem Difficulty
Crowds may be wise, but what happens when they encounter a very difficult problem (Krause et al., 2011)? Researchers in Germany tested the limits of “swarm intelligence” by asking visitors to a science exhibit to step up to a computer console and enter in their best guesses for two questions. One problem was easy: All they had to do was estimate the number of marbles in a jar next to the computer console. The second problem was more difficult: “Estimate how many times a coin needs to be tossed for the probability that the coin will show heads each time to be roughly as small as that of winning the German lotto” (p. 942). Statisticians who were also active lotto players may have known that it would take about 24 consecutive heads to equal the very remote chances of winning the lottery (1 in 35 million), but most visitors found the second question to be very challenging.
The crowd was quite wise when answering the easy question. The mean for the group was 553.6, coming within 1.5% of the actual number of marbles (562). But the crowd’s average estimate for the second question, 498 flips of the coin, was not accurate at all, suggesting a crowd will not be wise when its members lack the knowledge needed to solve the problem. These findings also affirmed the importance of recruiting enough people to form the crowd (see Figure 10.5). Even for the easy question (shown in the top chart), groups with ten members or fewer were not as accurate as the majority of the individual members. Groups that ranged in size from 10 to 40 members outperformed most of the individuals in the group, but not the top 25% of individuals. Once the group included more than 40 members, the group’s accuracy surged past most of the group’s members, but this effect occurred only for the relatively easy estimation problem. Estimates for the difficult problem, shown in the lower chart in Figure 10.5, were always inferior to individual member’s scores—because so many of the group members’ estimates were massively incorrect. Errors did not have a chance to cancel each other out. These and other findings suggest that compensatory decisional methods can often be improved by identifying the wiser ones in the crowd, and then weighting their estimates more heavily than the rank-and-file crowd members when aggregating the results (Budescu & Chen, 2014).
10-3d. Disjunctive Tasks
Sully, Skiles, and Harten had a decision to make. Without power, Flight 1549 needed a place to land. Harten, the air traffic controller, favored a return to LaGuardia. Skiles did not express any preference. Sully initially considered the nearby Teterboro airport, but then changed his choice to the Hudson. There could be only one solution—they could not switch to Plan B if Plan A failed.
When groups work at disjunctive tasks, they must generate a single solution that will stand as the group’s outcome. Juries making decisions about guilt or innocence, computer technicians deciding which program bug to fix first, or the coaching staff setting the lineup for the day’s game, are all performing disjunctive tasks. These types of tasks tend to be both unitary and optimizing, for they cannot be broken down into subtasks, and they require a high-quality or correct solution rather than a large quantity of product.
Disjunctive tasks often require discussion and decisions; Chapter 12 provides a more detailed analysis of how groups tackle such tasks. In general, however, groups perform disjunctive tasks better than most of the individual members. For example, if four students complete a quiz as a group, the group will likely outscore most of the individual students because more heads means more information and better detection of errors. In an actual class where students were permitted to take tests in pairs or alone, pairs scored nearly 4% points higher than individuals (Pandey & Kapitanoff, 2011).
When Does “Truth” Win?
When aviation experts reviewed all the facts related to Flight 1549, they concluded Sully’s solution was the best one given the circumstances. The flight crew’s choice of that alternative illustrates the truth-wins rule. The best solution (truth), when suggested during the group’s deliberations, was the one the group adopted (wins).
Truth, however, does not always win, for in some cases even though the correct answer is known by at least one group member, the group fails to select it as the group’s solution. Rosa may be certain that the answer to the question “Who first documented the reduction of individual productivity when in groups?” is “Ringelmann,” but her group may not accept her solution because they doubt her skills or because someone of higher status may propose a different solution. Ringelmann is the correct answer, but this truth will not win out over error unless someone in the group supports Rosa and her answer—a truth-supported-wins rule.
The truth-wins rule usually holds for groups working on Eureka problems, which are ones with obviously correct answers. When we are told the answer to a Eureka problem, the answer fits so well we react with an “Aha!” or “Eureka!” The answers to non-Eureka problems, in contrast, are not so satisfying, and so the truth-supported-wins rule holds for groups working on those types of problems. Even after arguing about them, we often wonder if the recommended answer is the correct one.
Consider, for example, the famous horse-trading problem:
A man bought a horse for $60 and sold it for $70. Then he bought it back for $80 and again sold it for $90. How much money did he make in the horse-trading business? (Maier & Solem, 1952, p. 281)
When 67 groups discussed this problem, many included a member who knew the correct answer, but even these groups often adopted the wrong solution. In this case, truth lost because knowledgeable members had a difficult time persuading the other members to adopt their solutions. In fact, some people later changed their answers to match the incorrect solution advocated by their groups (Maier & Solem, 1952; the answer, by the way, is $20). Thus, groups perform at the level of the best member of the group only if
(1)
the member who knows the answer shares his or her answer with the others and
(2)
the group decides to adopt this answer as the solution (Davis, 1973; Littlepage, 1991; Steiner, 1972).
Intellective and Judgmental Tasks
Groups are also more likely to recognize, and accept, the correct solution when the person who proposes it can demonstrate that the solution is the correct one. If a group member backs up a solution with a proof, a citation, or a quote from an authority, the rest of the group may accept it. But if he or she says, “It’s hard to explain, I just have a feeling that is the answer,” then even the correct answer may not find acceptance within the group.
10-3d. Disjunctive Tasks
Sully, Skiles, and Harten had a decision to make. Without power, Flight 1549 needed a place to land. Harten, the air traffic controller, favored a return to LaGuardia. Skiles did not express any preference. Sully initially considered the nearby Teterboro airport, but then changed his choice to the Hudson. There could be only one solution—they could not switch to Plan B if Plan A failed.
When groups work at disjunctive tasks, they must generate a single solution that will stand as the group’s outcome. Juries making decisions about guilt or innocence, computer technicians deciding which program bug to fix first, or the coaching staff setting the lineup for the day’s game, are all performing disjunctive tasks. These types of tasks tend to be both unitary and optimizing, for they cannot be broken down into subtasks, and they require a high-quality or correct solution rather than a large quantity of product.
Disjunctive tasks often require discussion and decisions; Chapter 12 provides a more detailed analysis of how groups tackle such tasks. In general, however, groups perform disjunctive tasks better than most of the individual members. For example, if four students complete a quiz as a group, the group will likely outscore most of the individual students because more heads means more information and better detection of errors. In an actual class where students were permitted to take tests in pairs or alone, pairs scored nearly 4% points higher than individuals (Pandey & Kapitanoff, 2011).
When Does “Truth” Win?
When aviation experts reviewed all the facts related to Flight 1549, they concluded Sully’s solution was the best one given the circumstances. The flight crew’s choice of that alternative illustrates the truth-wins rule. The best solution (truth), when suggested during the group’s deliberations, was the one the group adopted (wins).
Truth, however, does not always win, for in some cases even though the correct answer is known by at least one group member, the group fails to select it as the group’s solution. Rosa may be certain that the answer to the question “Who first documented the reduction of individual productivity when in groups?” is “Ringelmann,” but her group may not accept her solution because they doubt her skills or because someone of higher status may propose a different solution. Ringelmann is the correct answer, but this truth will not win out over error unless someone in the group supports Rosa and her answer—a truth-supported-wins rule.
The truth-wins rule usually holds for groups working on Eureka problems, which are ones with obviously correct answers. When we are told the answer to a Eureka problem, the answer fits so well we react with an “Aha!” or “Eureka!” The answers to non-Eureka problems, in contrast, are not so satisfying, and so the truth-supported-wins rule holds for groups working on those types of problems. Even after arguing about them, we often wonder if the recommended answer is the correct one.
Consider, for example, the famous horse-trading problem:
A man bought a horse for $60 and sold it for $70. Then he bought it back for $80 and again sold it for $90. How much money did he make in the horse-trading business? (Maier & Solem, 1952, p. 281)
When 67 groups discussed this problem, many included a member who knew the correct answer, but even these groups often adopted the wrong solution. In this case, truth lost because knowledgeable members had a difficult time persuading the other members to adopt their solutions. In fact, some people later changed their answers to match the incorrect solution advocated by their groups (Maier & Solem, 1952; the answer, by the way, is $20). Thus, groups perform at the level of the best member of the group only if
(1)
the member who knows the answer shares his or her answer with the others and
(2)
the group decides to adopt this answer as the solution (Davis, 1973; Littlepage, 1991; Steiner, 1972).
Intellective and Judgmental Tasks
Groups are also more likely to recognize, and accept, the correct solution when the person who proposes it can demonstrate that the solution is the correct one. If a group member backs up a solution with a proof, a citation, or a quote from an authority, the rest of the group may accept it. But if he or she says, “It’s hard to explain, I just have a feeling that is the answer,” then even the correct answer may not find acceptance within the group.
10-3e. Conjunctive Tasks
As Flight 1549 descended into the Hudson, the passengers sat in silence, bracing for the impact. But when Sully gave the order to evacuate the aircraft, the passengers filled the aisle, pushing to reach the exits. Some opened the overhead storage areas, seeking their carry-ons, but blocking the escape route. Some passengers, such as 85-year old Lucille Palmer, did not move as quickly as the other passengers. Many passengers seated at the back of the plane first tried the rear emergency exit, not knowing it was already underwater and could not be used. Sully stood by the door of the cockpit, waiting for the airplane to empty. The evacuation would not be complete, in his mind, until the last passenger and crew member had disembarked.
The Weakest Link
On most tasks, the group’s performance results from some combination of all the group members’ efforts. For conjunctive tasks, however, members are coupled together, like links in a chain, so interdependence of actions and outcomes is maximized. Mountain climbers linked by safety lines, pallbearers carrying a casket, and passengers exiting a flight are performing a conjunctive task, for the task is not completed until the last link in the chain finishes. Such groups sometimes perform exceptionally well, particularly when one member of the team sets a high standard for performance, and the other members are motivated to match that level of proficiency. However, with conjunctive tasks, the proficiency of the best member does not define the group’s performance. Instead, performance is determined by the proverbial “weakest link”: the slowest, least productive, least skilled, most ineffective member. The speed of a group of mountain climbers moving up the slope is determined by its slowest member. The trucks in a convoy can move no faster than the slowest vehicle. Had anyone died on Flight 1549 it would not have been described as a “miracle.”
Because of this coordination problem, groups often take steps to improve their proficiency on conjunctive tasks. If the conjunctive tasks are divisible, then the group can assign group members to the subcomponents that best match their skill levels. If the least competent member is matched with the easiest task, a more satisfying level of performance may be obtainable. If the least competent member is matched with a difficult subtask, group performance will, of course, decline still further. (See Steiner, 1972, Chapter 3, for a detailed review of group performance on divisible tasks.)
The Köhler Effect
Few group members relish being cast in the role of the group’s most inferior group member (IGM), so they often respond to this indignity by expending more effort than they would if they were working alone—a rare group motivation gain rather than loss. This tendency is known as the Köhler effect, named after Otto Köhler, the researcher who first documented the performance gains of weaker individuals striving to keep up with the accomplishments of others in the group (Köhler, 1926; Witte, 1989).
Social psychologist Norbert Kerr and his colleagues (2007) studied the Köhler effect by arranging for women to complete a simple weight-lifting task. They were told to hold a three-pound dumbbell horizontally for as long as they could. When they lowered the weight, it would break a trip wire monitored by a laboratory computer, and the trial would end. The longer they held the weight, the more money they could possibly earn at the end of the study. They completed this task four times, with both their dominant and nondominant arm. Women assigned to the control condition completed the task without any partner; they thought they were alone. Others, however, were led to believe that an “Anne Roberts” was in the next room and that she was also performing the task. In both the coaction and conjunctive conditions, participants could monitor Anne’s performance on Trials 3 and 4 via computer as they themselves struggled to hold up their weight. But in the conjunctive condition, participants were also told that whoever lowered her weight first would determine the group’s score. Since Anne did not actually exist (and therefore never tired), subjects were always the IGM. But reluctant IGMs, judging by how much longer they managed to hold up the weight when paired with Anne; they achieved a 20-second gain in the coaction condition and a 33-second gain in the conjunctive condition.
A meta-analytic review of 22 studies of group performance confirms these findings. Individuals who find that their work is inferior to someone else’s work show improvement relative to others deprived of this comparison information, but this performance gain is particularly dramatic when they are part of a group working on a conjunctive task. IGMs are much more likely to improve when in face-to-face groups and when information about the quality of other people’s performance is readily available. The Köhler effect is also stronger in women than men (Weber & Hertel, 2007).
10-3f. Discretionary Tasks
Steiner noted that a group can sometimes complete the tasks it faces by using a variety of combination procedures. How, for example, would a group estimate the temperature of the room in which it is working? One simple method would involve averaging individual judgments. Alternatively, members can determine whether anyone in the group is particularly good at such judgments, and then they use this person’s answer as the group solution. Judging the temperature of the room is a discretionary task, because the members themselves can choose the method for combining individual inputs.
10-3g. Process Gains in Groups
Steiner’s (1972) analysis of task demands and their impact on group performance is summarized in Table 10.4. Groups perform additive tasks fairly well, although their productivity is often limited by social loafing. Groups also perform better than the average group member on many other kinds of tasks (compensatory, disjunctive, divisible conjunctive tasks when weaker members are assigned easier subtasks, and discretionary), but only when process losses are minimized. As Steiner’s (1972) formula,
AP
=
PP
-
PL
, predicts, process losses turn potential productivity into actual productivity.
Table 10.4 A Summary of the Potential Productivity of Groups Working on Various Tasks
Type of Task Productivity Effect
Additive Better than the best: The group exceeds the performance of even the best individual member.
Compensatory Better than most: The group exceeds the performance of a substantial number of the individual members.
Disjunctive Better than average and sometimes equal to the best: The group performs best if it accepts the most capable member’s input as the group solution; groups rarely perform better than the best member. (Process gains resulting in synergy are rare.)
Conjunctive: Unitary Equal to the worst: The group equals the performance of its least capable member.
Conjunctive: Divisible Better than the worst: Performance will be superior if subtasks are matched to members’ capabilities.
Discretionary Variable: Performance depends on the combination rules adopted by the group.
But don’t groups sometimes achieve results that surpass what Steiner’s theory predicts? Can’t group members, by collaborating on a shared task, sometimes gain new solutions, energy, and insights into old problems that they would never have achieved as individuals? Aren’t some groups greater than the sum of their parts? Does 1 + 1 + 1 sometimes equal 4 instead of 3?
Searching for Synergy
Group researchers have long sought definitive evidence of synergy in groups. Synergy occurs whenever the combined effect of two or more discrete systems is greater than the effect of these systems when they operate independently. Two drugs, for example, combine synergistically if their effects are greater when they are taken together rather than separately. In groups, if synergy occurs, the group as a whole performs better than what would be expected given the skills and abilities of its members. Synergy, as defined by group researchers, is not group-level energy or a heightened sense of connectedness among members, but a process gain generated by performance-enhancing group processes. Synergy is sometimes called an assembly bonus effect because “the group is able to achieve collectively something which could not have been achieved by any member working alone or by a combination of individual efforts” (Collins & Guetzkow, 1964, p. 58).
Strong and Weak Synergy
Social psychologist James Larson (2010) draws a distinction between weak and strong synergy. Imagine four students who, working separately, earn 70%, 80%, 80%, and 90% on a test; their average score would be 80%. If when they earn a score of 85% when they take the test in their learning team—a score above the 80% average and a better score than three of the four would have achieved working individually—then the group would demonstrate weak synergy. The group would be showing strong synergy, however, if it scored a 91% or higher—better than even the best member of the group.
Synergy eludes most groups. Steiner did not write his formula as
AP
=
PP
-
PL
+
PG
, where PG indicates process gains due to synergy. When individuals work on a collective task, the whole is often much less than the sum of the parts, as members exert less effort (social loafing) or let others do their share of the work (free-riding). Groups often outperform the most incompetent group member (the “better than the worst” effect), and, in most cases, they perform as well as the most typical group member. Rarely, however, do they perform above and beyond the level of the typical group member (weak synergy) or better than the best member (strong synergy; see Carey & Laughlin, 2012).
But synergy, although rare, does happen, particularly when the group members are highly motivated—when grades, jobs, or lives are on the line. Synergy happens, for example, when students work in learning teams in their classes, and each student’s grade in the course is based, in part, on their group’s collective performance. These groups often outperformed their best members, suggesting that the groups could identify new and better solutions when they worked together (Michaelsen, Watson, & Black, 1989). Other investigators, who replicated these findings, concluded that the synergistic effects occurred primarily because someone in the group other than the best member knew the right answer and could correct the best member (Stasson & Bradshaw, 1995). Such groups apparently make this critically important judgment by considering the level of confidence each member expresses in his or her answer. The group is more likely to perform well when the member who is correct is also the most confident (Bahrami et al., 2010; see too Koriat, 2012).
A group’s chances of achieving strong synergy also depend on the basal skill level of the group members and the “cognitive distance” separating them. Groups with highly competent members, given their greater resources, are more likely to display strong synergy than are groups whose members are less skilled or less prepared (Curşeu, Jansen, & Chappin 2013). Synergy also requires some, but not too much, variation among the members in cognitive competency and expertise. If all the members are nearly equivalent in their level of competence, their resources may be so redundant that little is gained by pooling those resources. Increasing cognitive distance therefore increases the group’s chances to achieve strong synergy, except when the most competent group member is too advanced. In such cases, the group has little to offer the highly competent member, who in turn may have trouble convincing the other group members to accept his or her recommendations (Curşeu et al., 2014).
Groups can, however, improve their pursuit of synergy by deliberately adopting specialized performance methods that are designed to minimize all process loss, while maximizing the possibility of achieving process gain. The final section of this chapter examines several of these methods, particularly those that can be used by groups searching for creative solutions to difficult problems.
10-4. Group Creativity
Sully and Skiles did not have much time to troubleshoot the problem and discuss alternatives. Their air traffic controller had proposed a few alternatives. They tried to restart the engines, but as the plane lost altitude, Sully turned to Skiles and asked, “Got any ideas?” Their final solution, however, if not creative was certainly unusual. Commercial jets rarely land in the Hudson River.
10-4a. Brainstorming
Had the crew of Flight 1549 more time they may have used a group method known as brainstorming to identify solutions to their problem. This method was developed by Alex Osborn (1957), an advertising executive, to help his colleagues identify novel, unusual, and imaginative solutions. The technique requires an open discussion of ideas and is guided by four basic rules:
Be expressive. Express any idea that comes to mind, no matter how strange, wild, or fanciful. Do not be constrained or timid; freewheel whenever possible.
Postpone evaluation. Do not evaluate any of the ideas in any way during the idea-generation phase. All ideas are valuable.
Seek quantity. The more ideas, the better. Quantity is desired, for it increases the possibility of finding an excellent solution.
Piggyback ideas. Because all ideas belong to the group, members should try to modify and extend others’ ideas whenever possible. Brainstorming is conducted in a group, so that participants can draw from one another.
Does Brainstorming Work?
When groups need to think of new ideas, the call to “brainstorm” is often raised, but their faith in this method may be misplaced. Researchers began testing this method by comparing brainstorming groups to individuals and to so-called nominal groups: groups created by having individuals work alone and then pooling their ideas (a group “in name” only). Their studies offered support for brainstorming. A four-person brainstorming group, for example, would not only outperform any single individual but also a nominal group of four individuals. However, these investigations stacked the deck against the nominal groups; brainstorming groups were told to follow the four basic brainstorming rules, whereas the individuals composing the nominal group were not given any special rules concerning creativity. When individuals working alone were better informed about the purposes of the study and the need for highly creative responses, they often offered more solutions than individuals working in groups. In one study, for example, four-person groups came up with an average of 28 ideas in their session, whereas four individuals working alone suggested an average of 74.5 ideas when their ideas were pooled. The quality of ideas was also lower in groups—when the researchers rated each idea on creativity, they found that individuals had 79.2% of the good ideas. Groups also performed more poorly even when given more time to complete the task (Diehl & Stroebe, 1987; see Paulus & Brown, 2007; Paulus & Coskun, 2013 for reviews).
Production Blocking
Brainstorming groups, like many performing groups, must struggle to overcome process losses as they strive to generate ideas. Even though members think they are expending maximum effort, social loafing detracts from their performance unless such safeguards as high identifiability, clear goals, and involvement prevent the undercutting of individual effort (Wegge & Haslam, 2005). But brainstorming groups also suffer coordination and cognitive losses. The originators of brainstorming thought that hearing others’ ideas would stimulate the flow of ideas, but the clamor of creative voices instead can cause production blocking. In brainstorming groups, members must wait their turn to get the floor and express their ideas, and, during that wait, they forget their ideas or decide not to express them. Hearing others is also distracting and can interfere with one’s ability to do the cognitive work needed to generate ideas. Even when researchers tried to undo this blocking effect by giving brainstormers notepads and organizing their speaking turns, the groups still did not perform as well as individuals who were generating ideas alone (Diehl & Stroebe, 1987, 1991; Nijstad & Stroebe, 2006).
Evaluation Apprehension
Evaluation apprehension can also limit the effectiveness of brainstorming groups, even though the “no evaluation” rule was designed to free members from such concerns (Diehl & Stroebe, 1987). Groups become even less effective when an authority watches them work. Apparently, members worry that the authority may view their ideas negatively (Mullen et al., 1991). Individuals with high social anxiety are particularly unproductive brainstormers and report feeling more nervous, anxious, and worried than group members who are less anxiety prone (Camacho & Paulus, 1995).
Social Matching
Social comparison processes also conspire to create a social matching effect. Although undercontributors are challenged to reach the pace established by others, overcontributors tend to reduce their contributions to match the group’s mediocre standards. Since overcontribution is more effortful than undercontribution, over time the high performers tend to adjust their rate downward to match the group’s lower norm (Brown & Paulus, 1996; Seta, Seta, & Donaldson, 1991).
Illusion of Group Productivity
Brainstorming groups are also unproductive because they often overestimate their productivity. In many cases, a group has no standard to determine how well it is performing, so individual members can only guess at the quantity and quality of their group’s product and their personal contributions to the endeavor. These estimates, however, are often unrealistically positive, resulting in a robust illusion of group productivity (Stroebe, Diehl, & Abakoumkin, 1992). Members of groups working on collective tasks generally think that their group is more productive than most (Polzer, Kramer, & Neale, 1997). Nor do group members feel that they are doing less than their fair share. When members of a group trying to generate solutions to a problem were asked to estimate how many ideas they provided, each group member claimed an average of 36% of the ideas, when in reality they generated about 25% of the ideas (Paulus et al., 1993).
Several processes appear to combine to sustain this error in performance appraisal. Group members may intuitively mistake others’ ideas for their own, and so, when they think about their own performance, they cognitively claim a few ideas that others actually suggested (Stroebe et al., 1992). When they brainstorm in groups, they can also compare themselves to others who generate relatively few ideas, reassuring them that they are one of the high performers (Paulus et al., 1993). Group brainstorming may also “feel” more successful since the communal process means that participants rarely experience failure. When alone and trying to think creatively, people repeatedly find that they are unable to come up with a new idea. In groups, because others’ ideas are being discussed, people are less likely to experience this failure in their search for new ideas (Nijstad, Stroebe, & Lodewijkx, 2006). Group members also mistakenly adjust their definition of what counts as a group success downward when they brainstorm in groups. If they estimate that a single person can generate 10 good ideas, do they think a group of 5 people will come up with 50 good ideas? No. They expect fewer contributions per member, and their expectations decline still further the larger the group (Jones & Lambertus, 2014).
10-4a. Brainstorming
Had the crew of Flight 1549 more time they may have used a group method known as brainstorming to identify solutions to their problem. This method was developed by Alex Osborn (1957), an advertising executive, to help his colleagues identify novel, unusual, and imaginative solutions. The technique requires an open discussion of ideas and is guided by four basic rules:
Be expressive. Express any idea that comes to mind, no matter how strange, wild, or fanciful. Do not be constrained or timid; freewheel whenever possible.
Postpone evaluation. Do not evaluate any of the ideas in any way during the idea-generation phase. All ideas are valuable.
Seek quantity. The more ideas, the better. Quantity is desired, for it increases the possibility of finding an excellent solution.
Piggyback ideas. Because all ideas belong to the group, members should try to modify and extend others’ ideas whenever possible. Brainstorming is conducted in a group, so that participants can draw from one another.
Does Brainstorming Work?
When groups need to think of new ideas, the call to “brainstorm” is often raised, but their faith in this method may be misplaced. Researchers began testing this method by comparing brainstorming groups to individuals and to so-called nominal groups: groups created by having individuals work alone and then pooling their ideas (a group “in name” only). Their studies offered support for brainstorming. A four-person brainstorming group, for example, would not only outperform any single individual but also a nominal group of four individuals. However, these investigations stacked the deck against the nominal groups; brainstorming groups were told to follow the four basic brainstorming rules, whereas the individuals composing the nominal group were not given any special rules concerning creativity. When individuals working alone were better informed about the purposes of the study and the need for highly creative responses, they often offered more solutions than individuals working in groups. In one study, for example, four-person groups came up with an average of 28 ideas in their session, whereas four individuals working alone suggested an average of 74.5 ideas when their ideas were pooled. The quality of ideas was also lower in groups—when the researchers rated each idea on creativity, they found that individuals had 79.2% of the good ideas. Groups also performed more poorly even when given more time to complete the task (Diehl & Stroebe, 1987; see Paulus & Brown, 2007; Paulus & Coskun, 2013 for reviews).
Production Blocking
Brainstorming groups, like many performing groups, must struggle to overcome process losses as they strive to generate ideas. Even though members think they are expending maximum effort, social loafing detracts from their performance unless such safeguards as high identifiability, clear goals, and involvement prevent the undercutting of individual effort (Wegge & Haslam, 2005). But brainstorming groups also suffer coordination and cognitive losses. The originators of brainstorming thought that hearing others’ ideas would stimulate the flow of ideas, but the clamor of creative voices instead can cause production blocking. In brainstorming groups, members must wait their turn to get the floor and express their ideas, and, during that wait, they forget their ideas or decide not to express them. Hearing others is also distracting and can interfere with one’s ability to do the cognitive work needed to generate ideas. Even when researchers tried to undo this blocking effect by giving brainstormers notepads and organizing their speaking turns, the groups still did not perform as well as individuals who were generating ideas alone (Diehl & Stroebe, 1987, 1991; Nijstad & Stroebe, 2006).
Evaluation Apprehension
Evaluation apprehension can also limit the effectiveness of brainstorming groups, even though the “no evaluation” rule was designed to free members from such concerns (Diehl & Stroebe, 1987). Groups become even less effective when an authority watches them work. Apparently, members worry that the authority may view their ideas negatively (Mullen et al., 1991). Individuals with high social anxiety are particularly unproductive brainstormers and report feeling more nervous, anxious, and worried than group members who are less anxiety prone (Camacho & Paulus, 1995).
Social Matching
Social comparison processes also conspire to create a social matching effect. Although undercontributors are challenged to reach the pace established by others, overcontributors tend to reduce their contributions to match the group’s mediocre standards. Since overcontribution is more effortful than undercontribution, over time the high performers tend to adjust their rate downward to match the group’s lower norm (Brown & Paulus, 1996; Seta, Seta, & Donaldson, 1991).
Illusion of Group Productivity
Brainstorming groups are also unproductive because they often overestimate their productivity. In many cases, a group has no standard to determine how well it is performing, so individual members can only guess at the quantity and quality of their group’s product and their personal contributions to the endeavor. These estimates, however, are often unrealistically positive, resulting in a robust illusion of group productivity (Stroebe, Diehl, & Abakoumkin, 1992). Members of groups working on collective tasks generally think that their group is more productive than most (Polzer, Kramer, & Neale, 1997). Nor do group members feel that they are doing less than their fair share. When members of a group trying to generate solutions to a problem were asked to estimate how many ideas they provided, each group member claimed an average of 36% of the ideas, when in reality they generated about 25% of the ideas (Paulus et al., 1993).
Several processes appear to combine to sustain this error in performance appraisal. Group members may intuitively mistake others’ ideas for their own, and so, when they think about their own performance, they cognitively claim a few ideas that others actually suggested (Stroebe et al., 1992). When they brainstorm in groups, they can also compare themselves to others who generate relatively few ideas, reassuring them that they are one of the high performers (Paulus et al., 1993). Group brainstorming may also “feel” more successful since the communal process means that participants rarely experience failure. When alone and trying to think creatively, people repeatedly find that they are unable to come up with a new idea. In groups, because others’ ideas are being discussed, people are less likely to experience this failure in their search for new ideas (Nijstad, Stroebe, & Lodewijkx, 2006). Group members also mistakenly adjust their definition of what counts as a group success downward when they brainstorm in groups. If they estimate that a single person can generate 10 good ideas, do they think a group of 5 people will come up with 50 good ideas? No. They expect fewer contributions per member, and their expectations decline still further the larger the group (Jones & Lambertus, 2014).
10-4c. Alternatives to Brainstorming
Most groups, electronic brainstorming (EBS) when faced with the challenge of generating creative solutions, uncreatively suggest brainstorming. But given the difficulties in implementing brainstorming techniques correctly, groups should consider turning to other methods in their quest for fresh ideas and new insights into old problems (Sunwolf, 2002).
The Nominal Group Technique
Several creativity-building methods take advantage of the “wisdom” of groups by integrating individual idea-generating sessions with group-level methods. The nominal group technique (NGT), for example, minimizes blocking and loafing by reducing interdependence among members; it achieves this improvement by starting with a nominal group phase before turning to a group session (Delbecq & Van de Ven, 1971).
Step 1. The group discussion leader introduces the problem or issue in a short statement that is written on a blackboard or flip chart. Once members understand the statement, they silently write ideas concerning the issue, usually working for 10–15 minutes.
Step 2. The members share their ideas with one another in a round-robin; each person states an idea, which is given an identification letter and written beneath the issue statement, and the next individual then adds his or her contribution.
Step 3. The group discusses each item, focusing primarily on clarification.
Step 4. The members rank the five solutions they most prefer, writing their choices on an index card.
The leader then collects the cards, averages the rankings to yield a group decision, and informs the group of the outcome. The group may wish to add two steps to further improve the procedure: a short discussion of the vote (optional step 5) and a revoting (optional step 6). These methods are particularly useful when groups discuss issues that tend to elicit highly emotional arguments. NGT groups produce more ideas and also report feeling more satisfied with the process than unstructured groups. The ranking and voting procedures also provide for an explicit mathematical solution that fairly weights all members’ inputs and provides a balance between task concerns and interpersonal forces (Delbecq & Van de Ven, 1971; Gustafson et al., 1973).
The Delphi Technique
The Delphi technique eliminates the group-level discussion altogether. This method, named for the legendary Delphic oracle, involves surveying members repeatedly with the results of each round of surveys informing the framing of the questions for subsequent rounds. The Delphi coordinator begins the process by developing a short list of questions on the topic and gathering the answers of a carefully selected group of respondents. Their answers are then pooled and communicated back to the entire group; members are asked to restate their responses to the original items, comment on others’ responses, or respond to new questions that emerged in the first round of surveying. This process is repeated until a solution is reached. The method is particularly well-suited for problems that cannot be solved by a systematic review of the available data (Forsyth, 2010).
Buzz Groups, Bug Lists, and Beyond
When stumped for new ideas, members can break up into buzz groups, which are small subgroups that generate ideas that can later be discussed by the entire group. Members can jot down a bug list of small irritations pertaining to the problem under discussion, and the group can then discuss solutions for each bug. Groups can use the stepladder technique, which requires asking each new member of the group to state his or her ideas before listening to the group’s position (Rogelberg & O’Connor, 1998). Groups can even use elaborate systems of idea generation with such exotic-sounding names as synectics and TRIZ. In synectics, a trained leader guides the group through a discussion of members’ goals, wishes, and frustrations using analogies, metaphors, and fantasy (Bouchard, 1972). TRIZ is used primarily in science and engineering and involves following a specific sequence of problem analysis, resource review, goal setting, and review of prior approaches to the problem (Moehrle, 2005).
Resources
Chapter Case: Miracle on the Hudson
Highest Duty by Chesley “Sully” Sullenberger (with Jeffrey Zaslow, 2009), the autobiographical account of the crash of Flight 1549, provides critical details about the processes that occurred on the flight deck before and after the bird collision.
Miracle on the Hudson by the Survivors (with William Prochnau and Laura Parker, 2010) provides details about what occurred in the passenger area of Flight 1549, drawn from the passengers’ personal statements about the experience.
Group Productivity
“Performance” by Bernard A. Nijstad (2013) is an excellent overview of group performance and productivity, but those seeking even more information about working groups may wish to consider Nijstad’s (2009) book, Group Performance.
In Search of Synergy in Small Group Performance by James R. Larson, Jr. (2010) examines synergy in groups generating ideas, solving problems, rendering decisions, and making judgments.
Social Facilitation and Loafing
“Social Facilitation: Using the Molecular to Inform the Molar” by Allison E. Seitchik, Adam J. Brown, and Stephen G. Harkins (2016) examines empirical findings pertaining to improved performance in the presence of others, including studies examining the threat-induced potentiation of prepotent responses model of social facilitation.
“Understanding Individual Motivation in Groups: The Collective Effort Model” by Steven J. Karau and Kipling D. Williams (2001) is an updated review of work examining the factors that contribute to motivation loss in groups. This chapter is one of many excellent papers in Groups at Work, edited by Marlene E. Turner (2001).
Group Creativity
“Getting the Most Out of Brainstorming Groups” by Paul B. Paulus, Jubilee Dickson, Runa Korde, Ravit Cohen-Meitar, and Abraham Carmeli (2016) organizes much of the research on brainstorming and offers recommendations for eliminating impediments to creativity in groups.
Group Genius: The Creative Power of Collaboration by Keith Sawyer (2007) provides dozens of inspiring examples of groups that reached the heights of creativity through collaboration.
Chapter Case: Miracle on the Hudson
Highest Duty by Chesley “Sully” Sullenberger (with Jeffrey Zaslow, 2009), the autobiographical account of the crash of Flight 1549, provides critical details about the processes that occurred on the flight deck before and after the bird collision.
Miracle on the Hudson by the Survivors (with William Prochnau and Laura Parker, 2010) provides details about what occurred in the passenger area of Flight 1549, drawn from the passengers’ personal statements about the experience.
Group Productivity
“Performance” by Bernard A. Nijstad (2013) is an excellent overview of group performance and productivity, but those seeking even more information about working groups may wish to consider Nijstad’s (2009) book, Group Performance.
In Search of Synergy in Small Group Performance by James R. Larson, Jr. (2010) examines synergy in groups generating ideas, solving problems, rendering decisions, and making judgments.
Social Facilitation and Loafing
“Social Facilitation: Using the Molecular to Inform the Molar” by Allison E. Seitchik, Adam J. Brown, and Stephen G. Harkins (2016) examines empirical findings pertaining to improved performance in the presence of others, including studies examining the threat-induced potentiation of prepotent responses model of social facilitation.
“Understanding Individual Motivation in Groups: The Collective Effort Model” by Steven J. Karau and Kipling D. Williams (2001) is an updated review of work examining the factors that contribute to motivation loss in groups. This chapter is one of many excellent papers in Groups at Work, edited by Marlene E. Turner (2001).
Group Creativity
“Getting the Most Out of Brainstorming Groups” by Paul B. Paulus, Jubilee Dickson, Runa Korde, Ravit Cohen-Meitar, and Abraham Carmeli (2016) organizes much of the research on brainstorming and offers recommendations for eliminating impediments to creativity in groups.
Group Genius: The Creative Power of Collaboration by Keith Sawyer (2007) provides dozens of inspiring examples of groups that reached the heights of creativity through collaboration.