Note that this approach assumes that the outcomes of the binary payoff distribution are, by one means or another, known to the decision maker, an issue we will return to in the Discussion. Based on this idea and under the additional assumption of parameter point inference, the definition of the remaining components of the MEU framework for application to the SSP is straightforward:. Likewise, we define the set of possible acts, i. Intuitively, no information inherent in a sample of size n that is relevant for the parameters of the underlying probabilistic model is lost when recording the number of 1's r n and the total number of samples n , rather than the complete sequences of ones and zeros.
In summary, we define the problem space as:. Below, we will choose a simple, familiar form for the terminal utility. As in 13 we set. Finally, we denote the experiment-dependent probability measure on the space of possible true states of the world and experimental outcomes by:. We then choose the quadratic loss function as a familiar example of an opportunity loss function. As discussed in Pratt et al. In other words, a terminal utility function is expressed in terms of a terminal opportunity loss function:.
The solution in terms of the optimal action is identical. The expected opportunity loss associated with a given act, however, corresponds to the negative of the expected terminal utility associated with the same act shifted by an additive constant, which is independent of a. To determine the value of the expected terminal utility function based on the minimization of the expected terminal opportunity loss function, however, the decision maker requires the evaluation of the second term in Put succinctly: if the decision maker is merely interested in choosing the optimal action, it may formulate the problem either in terms of terminal opportunity loss or in terms of terminal utility.
Only if, in addition, the decision maker is interested in the associated utility of the optimal act, the decision maker is required to formulate the problem in terms of a utility function. A different perspective on the additive constant in 22 is afforded by considering the terminal utility of no experimentation. If the decision maker chooses not to experiment at all, i.
This corresponds to the integration of the first term in 22 under the prior distribution and its maximization. Because of this normalization property, and because we have introduced the MEU framework in terms of a terminal utility function, which is to be maximized, we will maintain this convention and explicitly take the second term in 22 into account. The quadratic terminal loss function is zero, if the chosen act, i. Summarizing 22 and 23 , we thus define the following terminal utility function in the current section.
The beta prior distribution in the Binomial sampling scenario has well-known advantages: It is specified on the parameter space of interest for Binomial sampling, it is the conjugate-prior distribution for the Binomial distribution, i. Additionally, we depict the dependence of the beta distribution expectation and standard deviation on its parameters in Figure 3B.
Figure 3. The first row depicts the joint distribution for the reference prior parameter settings, the second row for a prior distribution centered on 0. B Dependence of the first two central moments of the beta distribution on its parameters. Note that we have exchanged maximization with minimization for simplicity. As shown by Pratt et al.
Based on 30 one finds that the inner integral expression in the first term of 29 evaluates to:. Figure 4. The higher the prior certainty i. Notably, the minimized posterior terminal opportunity loss is symmetric in the prior parameters and decreases with higher prior certainty. Note that the posterior terminal opportunity loss is a function of the experimental outcome.
Substitution of the optimal posterior act obviously fulfills the minimization operation in 29 and integration with respect to the marginal outcome distribution then yields the complete integral term in 29 as:. For a proof of this identity, please see the Supplementary Material. Notably, the expected minimal terminal opportunity loss is monotonically decreasing with sample size and approaches zero for large n and is dependent on the choice of the prior distribution. From 35 we see that the optimal sample size for the SSP in the current scenario of a quadratic loss function and a beta prior distribution is a function of the prior distribution parameters and the sampling cost constant, which in analogy to 12 , we may write as:.
Analytical minimization of this function see Supplementary Material yields the explicit closed-form solution for the optimal sample size in the current scenario as. In Figure 5A , we visualize the expected terminal utility, the function h , and its maximum for three different settings of prior parameters.
In Figure 5B we visualize the optimal sample size as a function of the prior distribution parameters for two different settings of the sampling cost constant c. Notably, the sampling cost constant scales the optimal sample size, but does not affect the functional relationship of optimal sample size and prior distribution [which is, of course, also apparent from 37 ]. Figure 5. Optimal sample sizes for parameter point estimation. Higher sampling cost implicates lower optimal sample sizes, while the optimal sample size dependence on the prior parameter size is not affected.
Now, we assume that the decision maker is willing to accept a degree of uncertainty about the true state of the world upon selection of the optimal action. To this end, following Bernardo , we assume that the action space corresponds to the space Q of probability distributions over the state space S , or more formally:. More specifically, under the assumption of the action space corresponding to the space of probability distributions over states, the fundamental problem in determining an optimal sample size can be framed as a trade-off between maximizing the well-defined information about the parameter and minimizing sample cost.
Finally, we employ the same prior distribution as used above and evaluate the resulting optimal sample sizes. Continuing from the introduction of the terminal utility function u t in Equation 5 , we consider the consequences of identifying the action space with the space Q of probability distributions on the state space. Denoting the members of this space by q s , the terminal utility function takes the form:.
James, William | Internet Encyclopedia of Philosophy
This condition ensures that the decision maker chooses an action [i. While the requirements for properness and locality are well-motivated on the background of inferential decision problems for example in scientific contexts as they, for example, foster honesty , for the application to the SSP, the consequences, rather than the preconditions, of using logarithmic score functions are perhaps more important.
It is this last quantity, which the decision maker can trade-off with the expected cost of sampling in order to come to a judgment of the optimal sample size. Because the only argument dependent on q s on the right hand side of the above is u t q s , s , we may rewrite the above as:. By definition, the logarithmic score function is maximized for the posterior distribution p e s z and we obtain:. If one additionally defines the coefficient a to be 1 and, and in analogy to the notion of terminal opportunity loss above, the terminal utility of reporting the prior distribution over states p s to be zero Bernardo, one has:.
The quantity 46 is the well-known Kullback-Leibler divergence Kullback and Leibler, between the posterior distribution p n s z and the prior distribution p n s. The Kullback-Leibler divergence is a versatile quantity with ubiquitous applications in probabilistic modeling and here appears in form of a terminal utility as a consequence of the MEU framework with logarithmic score functions.
As noted by Lindley this quantity may alternatively be expressed as the experiment-dependent mutual information between the random variable representing possible states of the world s and the random variable representing experimental outcomes z , i.
Text Psyche Mining: Normative Vision: (3rd Edition)
Equation 49 captures the intuition that the task of selecting a good sample size or, more generally, designing a good experiment corresponds to maximizing the information that potential outcomes contain about the underlying state of the world [which, for real experiments has to be traded off against the cost of experimentation as apparent from 47 ].
We next apply these general results to the SSP. As discussed above, we assume A to be the space of probability distributions on [0, 1], here represented by probability density functions:. Note that the assumption of the elements of Q being probability density functions is a mere notational convenience. The results could equally well be formulated in terms of probability mass functions defined on suitably chosen partitions of [0, 1] see Bernardo and Smith, In summary, we thus assume the MEU problem space:.
The form of the utility function was discussed in detail above. Further, as above we assume the following probability measure on the Cartesian product of the space of states of the world and experimental outcomes:. To obtain the optimal sample size, we now consider Equation 47 , which, based on the specifications above, evaluates to:.
In this equation, the integral term mirrors the minimized posterior terminal opportunity loss of Equation 29 and intuitively corresponds to the expected KL-divergence between the posterior and prior beta distributions under the marginal distribution of the data. The KL-divergence between two beta distributions is well-known Liu et al.
For a proof, please refer to the Supplementary Material. Unfortunately, unlike the corresponding function in the case of parameter point inference, the function h in 55 is not readily maximized analytically. Mirroring Figure 5 for the case of parameter point inference, in Figure 6 we depict the minimized posterior terminal utility of Equation 55 , the function h of Equation 55 , and optimal sample sizes under the assumption of Bayesian parameter inference for different values of the sampling cost constant c.
Figure 6. Optimal sample size for Bayesian parameter inference. B For two cost constants, the panels depict the optimal sample sizes as a function of the prior parameters. In our application of the MEU framework for parameter point or interval probability estimation we have so far assumed analytically tractable parameter prior probability distributions and terminal utility functions mostly for mathematical convenience.
However, the MEU framework is by no means limited to these special classes of probability distributions and terminal utility functions. In this Section we demonstrate how the optimal sample size for the inference approach to the SSP can be derived with the help of a computer for arbitrary, but numerically evaluable, prior distributions and terminal utility function. As we elaborate below, this approach is of particular relevance for applications of the theory developed here in an experimental context.
Note that our demonstration merely serves as a proof-of-principle and does not aim for the systematic evaluation of the errors introduced by the numerical approximation of analytic quantities or attempts to provide an in any way exhaustive coverage of possible prior and terminal utility functions. This approach is visualized in Figure 7. Figure 7. Numerical Bayesian inference for a Binomial likelihood function. For a discretization of the state space into 10 equally spaced bins centered at 0.
Note the difference in scale between the first and second panel. We further note that the integration operations can, for finite discrete state and outcome spaces and probability mass functions defined over these spaces, be evaluated by means of scalar products. Finally, the respective maximization operations can be evaluated using standard list sorting techniques available in numerical computing.
Here, some numerical error is introduced by the discretization of state space. However, the larger errors are introduced by the treatment of the sample size as a continuous variable in the analytical case. Figure 8. Numerical replication of optimal sample sizes for parameter point estimation with squared error loss function and beta prior distribution.
As a proof-of-principle that the MEU framework can yield an optimal sample size for arbitrary prior distributions and terminal utility function, we consider prior probability mass and terminal utility functions for a discretization of the state space into 10 equally spaced bins Figure 9. Specifically, we specify the following prior distribution over states Figure 9A.
Figure 9. Numerical proof-of-principle. For a detailed discussion of this figure, please refer to the main text. Leaving technicalities aside, we next elaborate on the applicability of the numerical solutions discussed above in a concrete experimental context. We address this scenario first from the perspective of the decision maker, i. Consider an experimental participant faced with the SSP.
In line with the inferential notion of our framework, we assume that the participant would like to solve the question of how many samples to draw before deciding whether to take a final draw with economic consequences by means of estimating the expected value of the SSP. For the current purposes, we assume that this has resulted in the distribution shown in Figure 9A. For the prior distribution of Figure 9A , this marginal distribution of experimental outcomes is shown in Figure 9B. Note that in the current framework this loss function has only hypothetical, subjective applications.
Note that this sampling cost can be conceived as the intrinsic or extrinsic time-constraint of the participant for carrying out the sampling—if more samples are drawn, the experimental procedure will take longer, and the participant may not want to sit in the experiment for the rest of her life. Having now specified all essential components, our framework implies that the experimental decision maker would like to determine the best sample size that allows her to maximize her subjective utility terminal utility and negative sampling cost.
Importantly, this concept of optimal behavior is completely subjective—from an objective viewpoint, there is no guarantee that, e. In other words, our framework does not address an objective or absolute form of optimality, but rather a subjective or relative form. Note, however, that this observed-frequency perspective, while illustrative, is not coherent with the interpretation of the participant's probabilistic model as a quantification of uncertainty, and that under this interpretation, one would content with the analytically determined expected value.
We next consider the experimental applicability of our numerical framework from the perspective of the experimenter. By considering the framework discussed here, we obviously assume that the experimenter is led by the intuition that the participant's prior assumptions about the state of the world and economic preferences are of importance when studying decision making under uncertainty. More specifically, an experimenter may view the framework discussed herein from two perspectives.
In case of the former, the question arises, how the participant's neurocognitive apparatus is able to implement or at least approximate the non-trivial computations involved. In case of the latter, the question arises, which cognitive processes may distort the mapping from prior beliefs and preferences to selected sample sizes—which in turn may lead to more psychological plausible accounts of the decision processes in the SSP.
Secondly, the experimenter may conceive the framework as a valid working hypothesis and, by fixing or inducing specific components of the framework, study others. Finally, by revealing the prior distribution, misestimation preference, and observed sample size, the experimenter can study the cost that the participant assigns to a single draw. Additional possibilities for using the current framework in the second way arise by experimentally inducing prior assumptions, terminal utility functions, and sampling costs and then observing the behavioral consequences in sampling behavior.
We further discuss the experimental value of the framework in the Discussion. In this study we have shown how a normative benchmark for optimal sample sizes in the DFE sampling paradigm can be developed based on results from classical statistical decision theory. More specifically, we have shown that assuming an inference approach to the sampling problem in DFE, optimal sample sizes are dependent on the desired inference type and can be quantitatively related to the decision maker's prior beliefs about the problem, the decision maker's value assigned to identifying the correct solution, and the decision maker's cost assigned to each sample.
We conclude with discussing the benefits and limitations of this framework for generating testable predictions and point to potential applications of the framework in experimental cognitive psychology. Perhaps the most fundamental benefit of the MEU framework in the context of DFE is that it is explicit and constructive: upon specification of the necessary concepts the state, action, experimental, and outcome spaces, the utility function, and the probability measure on the product of experimental and outcome space it will yield an optimal sample size.
From the perspective of behavioral experimentation this is helpful, because search behavior in DFE can be tested against quantitative predictions. Further, because of its generality, the MEU framework can be adapted to a wide range of conceivable utility functions for example those incorporating a notion of risk-sensitivity Shen et al.
Finally, because it is explicit with respect to its concepts and assumptions, it can serve as a reference point for more psychological plausible accounts of information search in DFE. For example, it may be argued that the assumption of a prior distribution over states and its ensuing observation-based update to a posterior distribution is a cognitively impossible task.
If the aim is to find a framework that does not require the specification of a joint probability measure on world states and experimental outcomes, but is still constructive insofar as it allows for the derivation of quantitative sample size predictions, then the MEU framework may serve as a starting point from which assumptions can successively be removed.
Analogous arguments can be made in order to infer participant's prior beliefs or subjective sampling costs. However, this is a technical issue, and its solution readily conceivable, even if not readily carried out. As an example, Daunizeau et al. Combining the current framework with suitable identifiability constraints and one of the mentioned scientific inference approaches thus allows for overcoming this limitation. Because of its indefiniteness, an unlimited set of objections can be raised against the current approach from the perspective of cognitive process modeling.
We thus limit ourselves to a set of objections for which we see constructive rejoinders at present. A first objection may be that the current application of the MEU framework assumes that participants have knowledge of the to be observed outcomes such that estimating the state of the world, i. We agree that this assumption has been made here for experimental approaches in DFE that work on a similar basis, see selected experiments in Erev et al. It should also be noted that in general, upon sampling, both outcomes will have been observed, permitting for the evaluation of the expected value estimate based on the inferred probability parameter values.
A second objection is that it is implausible that participants in DFE studies evaluate optimal sample sizes for each payoff distribution prior to starting the sampling. Instead, they may after each observation or sets thereof decide whether to a terminate the exploration phase and continue to the final incentivized draw, b continue sampling from the currently investigated payoff distribution, or c terminate sampling from the currently investigated payoff distribution and to start or continue to sample from another payoff distribution. A model class appropriate to capture these intuitions is offered by the theory of partially observable Markov decision processes Wiering and Otterlo, A valuable future contribution would be the explicit comparison of the numerical sample size predictions offered by the MEU and POMDP frameworks under identical prior, utility, and sampling cost assumptions.
Finally, we note that Vul et al. However, as Vul et al. At present, because of the fundamental difference of which distribution is being sampled, their approach is not easily related to the MEU framework and such an analytical treatment is beyond the scope of the current manuscript. Future analytical studies may shed light on the precise relationship between the MEU framework considered here and the work by Vul et al.
In summary, a broad empirical literature on DFE has developed over the last decade in behavioral psychology, which has shown that human choice behavior can remarkably differ depending on how information is presented and sampled for uncertain choices with economic consequences. However, so far few attempts have been made to study the quantitative nature of human sampling behavior in DFE by means of computational modeling. Specifically, we have shown how, under a probabilistic inference assumption, the optimal sample size in DFE can be quantitatively related to the decision maker's preferred type of inference, prior beliefs about the payoff distributions at hand, and utility assigned to the inference's precision.
Because of its quantitative nature, the framework introduced here has yielded directly testable predictions for the behavioral study of DFE. Moreover, given the strong conceptual similarity between the DFE sampling paradigm and evidence accumulation schemes as prevalent in research on perceptual decision making, we believe that the current study addresses key theoretical aspects of decision making under dynamic subjective uncertainty. Finally, we believe that the current study lays an important foundation for future theoretical efforts on the computational description of human behavior in the DFE sampling paradigm and provides a useful basis towards their experimental validation.
The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest. Barber, D. Bayesian Reasoning and Machine Learning. Google Scholar. Bernardo, J. Statistical inference as a decision problem: the choice of sample size. D 46, — Bayesian Theory. Bishop, C. New York, NY: Springer. Cover, T.
Elements of Information Theory. Hoboken, NJ: Wiley. Another example of conflict giving rise to emergent social identities is the conflict that has arisen over fracking of coal seam gas in Australia. Unlikely alliances have emerged between environmental group members, farmers, conservative politicians, and media presenters who oppose fracking, with government agencies and mining companies perceived as the salient outgroup Hutton, ; Colvin et al. There are important consequences of these identities that emerge out of environmental conflict.
The alignment of climate change attitudes with political party identity lends an intense and competitive intergroup dynamic to what should be even-handed discussions about science and truth. When framed within an entrenched intergroup context, solutions advanced by one political party are likely to be dismissed by political opponents simply because they emanate from the outgroup. To the extent that information comes from people, who are perceived to be aligned with the outgroup, ingroup members are more likely to dismiss it regardless of its veracity Abrams et al.
In this way intergroup distinctions become entrenched and the potential to reach compromise or develop viable solutions becomes less likely. Although intergroup conflict can stymy progress on environmental issues, it should be noted that a degree of intergroup conflict is inevitable when pushing for social change, and that the alternative to conflict is often an unhealthy stasis.
In their meta-analysis of the collective action literature, van Zomeren et al. Other research demonstrates that group identification is positively associated with the belief that the group can be effective in reaching its collective goals van Zomeren et al. Of course, reality is more complex than this: whether or not collective action readiness is viewed as positive and legitimate will likely depend on the type of collective action whether it is violent or non-violent; whether it involves trade-offs with economic goals, etc.
Furthermore, although environmental collective action often involves a range of people from many walks of life, research has shown that people hold negative stereotypes of environmentalists as militant, aggressive, unconventional, and eccentric Bashir et al. Bashir et al.
- Normative - Wikipedia.
- Philosophy of Economics;
- Flat Earthers: Belief, Skepticism, and Denialism | Psychology Today.
- When people reject facts, what do they really believe?.
With this in mind it is easy to see how messages that emanate from environmental groups that are perceived to be extreme may gain little traction with the broader populace and could even polarize people away from support for important environmental issues Bliuc et al. Two key factors influence which social identities guide behavior: fit and accessibility Oakes et al.
Comparative fit refers to the degree to which a social identity is seen to reflect real world differences between groups. Normative fit recognizes that categorization is a dynamic process that reflects the perceptions of perceivers; that is, people are more likely to categorize into ingroups and outgroups if differences between groups align with stereotypic expectations.
Social identities are also more or less likely to become the basis for self-definition depending on how accessible they are; some are fleetingly accessible if primed e. Some recent research demonstrates the fluid nature of social identities in the environmental domain. Rabinovich et al. In a similar vein, when students compared themselves to past students assumed to be less pro-environmental they judged current students to be more pro-environmental but when comparing current students with future students assumed to be more pro-environmental they judged current students to be less pro-environmental Ferguson et al.
Willingness to engage in sustainable behaviors also varied in line with the perceived ingroup norms, that is, there was greater willingness when participants compared with past students than when they compared with future students. These findings demonstrate how the intergroup comparative context can influence the content of social identity in ways that could facilitate or inhibit greater engagement in pro-environmental behavior and greater support for pro-environmental policy.
A question raised by our analysis is whether the social identity approach adds to the understanding of environmental problems beyond other prominent theoretical frameworks. In this section, we examine the similarities and differences between the social identity approach and other relevant theories, with a view to highlighting possibilities for integration and stimulating future directions across frameworks.
One of the most prominent theoretical lenses applied to understanding climate change beliefs is cultural cognition theory, which adapts the theorizing of Douglas and Wildavsky on cultural influences and risk perceptions. The theory of cultural cognition Kahan, ; Kahan et al. For example, people who subscribe to relatively individualistic and hierarchical values favor self-reliance, competition and free market solutions.
In contrast, people who subscribe to relatively communitarian and egalitarian values are concerned with social injustice, are suspicious of authority including industry and are committed to cooperation. Hence, those high on individualism and hierarchy are more inclined to value industry, downplay its risk to the environment and oppose regulation. Research guided by this theoretical framework has shown empirical evidence of the influence of these cultural values.
Specifically, when an expert presented climate change as a high risk, egalitarian communitarian participants were much more likely to agree that they were an expert than hierarchical individualist participants and vice versa, when the expert presented climate change as a low risk see also Price et al. What the cultural cognition and social identity approach share is the notion that people filter information through a particular lens—either through the lens of worldviews in the case of cultural cognition theory or through the lens of social identity and its associated norms.
Hence, both perspectives conclude that beliefs about climate change will depend on how climate change aligns with these important meaning-making psychological structures. Where the social identity approach departs from cultural cognition is in its focus on the context-dependent nature of identity. Worldviews and values are relatively static or at best slow to change whereas identity is fluid and may become more or less salient depending on the context. If we integrate across the two perspectives, cultural cognition theory suggests that people with individualistic and hierarchical values would be more likely to identify with conservative political parties.
If they do so and the identity becomes salient, then the norms of that identity will guide responses to issues that are group-relevant, such as climate change policy. This integration suggests that identity may mediate between cultural worldviews and environmentally related attitudes and behavior. One can also imagine, though, that some contexts may bring to the fore identities that would trump or at least attenuate cultural worldviews. For those who are less identified, their worldviews may have a greater influence on their attitudes to the workplace climate change policies. Future research that integrates across these two theories could test these hypotheses.
Within the environmental psychology literature there are two key theories that often frame research seeking to understand environmental decisions and behavior: the theory of planned behavior TPB; Ajzen, and value-belief-norm VBN theory Stern et al. The attraction of the TPB is that it is a parsimonious model—it proposes that attitudes, subjective norms, and perceived behavioral control predict intentions which in turn predict behavior.
Another advantage of the model is that its simplicity allows other relevant variables to be integrated into the model thereby increasing its predictive power Conner and Armitage, Social identity researchers have integrated social identity concepts into the TPB in two main ways: First, they have drawn on social identity theory to address the question of why subjective norms often emerge as the weakest predictor of intentions Armitage and Conner, From a social identity perspective, it is not necessarily the norms of important others in general that will predict environmental behavior intentions, but rather the norms of the most behaviorally relevant group Terry and Hogg, ; Terry et al.
Terry et al. Fielding et al. They showed that these intergroup perceptions emerged as an additional predictor above and beyond the TPB variables. Hence, the social identity approach complements the TPB and can increase its potential to understand and predict environmentally related behavior. It clarifies which norms are likely to influence behavior and highlights the potential of the intergroup context to influence environmental intentions. Another well-established approach to understanding environmentally significant individual behavior is the VBN theory proposed by Stern et al.
This theory proposes a causal sequence that moves from stable values and ecological worldviews to an awareness of consequences for the valued object e. At first glance the centrality of personal norms in the VBN runs counter to the primacy of social norms in the social identity perspective. However, the social identity approach conceives of the self as made up of social identity and personal identities and so these two approaches are not contradictory.
Stern also acknowledges that there are a range of factors that feed into environmental behavior, including features of the personal, social, and economic context, and that the influence of personal norms on behavior will depend on the importance of contextual factors. Where the influence of contextual factors are strong, attitudinal factors as outlined in the VBN will be relatively weak predictors of environmental behaviors. This conceptualization allows a comfortable co-existence between the social identity approach and the VBN—in some circumstances personal norms will be the main motivator of behavior whereas in others group-based social identity considerations will come to the fore.
Rather than seeing these two theoretical approaches as parallel processes, though, one could also imagine a feedback loop between social and personal identity. Being members of social groups that value the environment could lead ingroup members to internalize these group norms so that they become a strong personal norm.
Of course, it is also possible that individuals join groups on the basis of their values and so having environmentally oriented values predisposes people to joining groups that reinforce those values. Ultimately, longitudinal research is needed to disentangle the causal sequence. If we accept that the social identity approach offers a helpful theoretical lens through which to examine environmental attitudes and behavior, then it should also be able to offer solutions to address environmental problems and conflicts.
They are not exhaustive but are instead meant to provide a starting point that can stimulate future research to empirically test and further refine social identity approaches relating to the environment. Although some of these strategies have been discussed previously see e. Social identity strategies to encourage more pro-environmental attitudes and behaviors. When thinking about how to promote more positive pro-environmental outcomes, an important consideration is where the messages come from. Consistent with social identity theory we know that ingroup sources are perceived to be more trusted and credible and therefore more influential Hornsey et al.
This suggests the need for pro-environmental messages to come from ingroup members whenever possible. Of course, this may not always be possible; sometimes the environmental issue is a scientific or technical one requiring specific expertise. It may be possible even in this case for the outgroup spokesperson to emphasize a shared superordinate identity with the audience. As an example, Schultz and Fielding showed that when a scientist provided information about recycled water—a potentially contentious solution to address water shortage situations—and emphasized the social identity she shared with participants i.
Conflict between groups on environmental issues has the potential to impede progress on addressing these issues. As we outlined above, intergroup conflict reinforces the boundaries between groups so that group members relate more to each other as group members, and are therefore more likely to exhibit ingroup-favoring attitudes and behaviors. When thought of in this way it is easy to see how this type of intergroup context can stand in the way of developing bi-partisan climate change policy, or sustainable resource allocation that benefits the environment as well as other stakeholders.
One way that negative intergroup relations could be transformed is through forging a more inclusive superordinate identity that encompasses conflicting subgroups Gaertner et al. Past research has shown that this strategy can reduce prejudice and discrimination Gaertner and Dovidio, because outgroup members are now part of the ingroup and are therefore accorded the benefits of ingroup membership. Batalha and Reynolds highlight the importance of superordinate identity as a way to develop more effective global negotiations around climate change mitigation.
The ASPIRe model provides a process for forming superordinate identity beginning with identifying the current social identities that people use to define themselves, followed by the formation of subgroups and the articulation of subgroup goals. The final steps involve the overarching organizational group—incorporating all subgroups—formulating superordinate group goals that inform subsequent action. Samuelson et al. The formation of the San Antonio Watershed Council involved bringing a variety of stakeholders together in a structured communication setting that allowed collaborative learning.
The formation of the new group identity allowed stakeholders from different subgroups who came with opposing positions to reach consensus and develop a set of recommendations to improve the quality of the watershed. Forging a superordinate identity, though, should not entail group members losing or negating their subgroup identity. In fact, research has shown that there are greater reductions in intergroup bias when people identify with both their subgroup and a superordinate group simultaneously Hornsey and Hogg, a , b.
In the context of intergroup environmental conflicts, Opotow and Brook argue that retaining subgroup identities can allow the positive attitudes that can develop because of shared superordinate group identity to generalize to the broader subgroup. For example, in the context of the conflict relating to fracking of coal seam gas in Australia, the Lock the Gate Alliance includes farmers and environmentalists, groups that are usually not in alliance.
Preserving subgroup identities also reduces the risk that subgroup identities are threatened Hornsey and Hogg, a and allows an appreciation of the distinctiveness of subgroup identities such as the expertise and experiences of specific subgroups Opotow and Brook, A simple demonstration of the latter approach is provided in research by Van der Werff et al.
They show that reminding people of their past pro-environmental actions leads them to strengthen their identity as a pro-environmental person which subsequently leads to further pro-environmental behaviors. It is easy to envisage how this approach could be scaled up through simple messages that ask people to reflect on their various past actions that help to protect the environment.
Campaigns that address local environmental issues, such as drought, are also an opportunity to showcase pro-environmental norms i. Activating the regional identity could thereby make salient and strengthen the water conservation norms and could result in lower ongoing water consumption in the region. Following on from the point above, there is strong evidence that people are more likely to act in environmentally friendly ways when the norms of a behaviorally relevant ingroup—especially one that people identify highly with—are supportive of pro-environmental action.
It may not always be possible to make salient a social identity with supportive environmental norms and, as we noted previously, it is often the case that ingroups have positive environmental injunctive norms but negative environmental descriptive norms i. Research has shown that placing a greater emphasis on the injunctive norm can help to overcome the problem of a negative descriptive norm Schultz et al. Thus, rather than drawing attention to the negative ingroup descriptive norms e.
If ingroup norms are not pro-environmental, another strategy supported by the social identity approach and discussed above is to make salient a higher order social category that does have pro-environmental norms. For example, data show that young adults engage in less pro-environmental behavior than older age groups Eurobarometer, Hence, making a higher order identity salient—for example, a national identity that encompasses more pro-environmental age groups—may help to reinforce pro-environmental norms and positively influence pro-environmental behavior as these norms and behavior can be truthfully attributed to the broader ingroup.
This strategy may be more likely to be effective when the lower order category e. For example, communication that reminds people that members of their country support pro-environmental policy and behavior could be accompanied by images that incorporate a range of citizens including younger citizens. Whether or not this approach could work in contexts where the subordinate identity is particularly salient, such as young people making environmental decisions in the presence of their peers, remains an empirical question.
To this point, we have described strategies that could help to shift perceptions of ingroup norms. A more direct approach is suggested by Seyranian and Seyranian et al. This approach outlines a process whereby a leader can shift social identity content in a direction that can help to promote positive social change. They highlight the need for leaders to advance a vision to group members through the use of inclusive ingroup language e. As an example, when ingroup leaders advocated for renewable energy using inclusive language, support for renewable energy was perceived to be more ingroup normative and there were greater intentions on the part of ingroup members to act in relation to renewable energy Seyranian, Further research is needed to identify the critical elements that are most effective at changing the content of ingroup identity.
Throughout this paper, we have highlighted interesting and important questions that could be pursued by researchers within a social identity framework. In considering directions for future research, we encourage researchers adopting a social identity approach to focus on issues that can have significant environmental impact.
Full text issues
Psychological research in the environmental domain has been heavily weighted toward individual actions, such as recycling, or energy and water conservation. But there is a need to expand our focus and to pursue dependent variables that represent greater impact. In light of the lack of progress in instituting climate change and environmental policy in most countries, an important focus for social identity researchers should be on policy acceptance and providing communicators with the tools to convince people of the need for environmental protection policy.
From a social identity approach perspective, lack of policy acceptance reflects that environmental goals are not normative in many groups and this may arise in part because of intergroup conflicts that create divisions rather than bridges between groups. An important challenge for social identity researchers is to identify frames that can appeal to decision-makers and unite disparate groups who conflict over environmental issues.
For example, what is the best way to frame climate change policy that will elicit positive responses from political conservatives and liberals alike? Researchers have begun to recognize the importance of finding frames that appeal to differing values and ideologies, for example, Feygina et al. Focusing on frames that appeal to decision-makers and elites may be particularly important given their power to appeal to the broader group.
At present there is little research to provide evidence for what works and what does not. Effective collective action has the potential to sway segments of the community who do not currently have an opinion or stake in an issue and to send messages to elites and decision-makers. As we noted previously, SIMCA has clearly demonstrated that social identity is a key predictor of collective action and that the more politicized the identity, the stronger the relationship van Zomeren et al.
Traditional forms of collective action have centered on taking part in protests or rallies. Marching alongside people who share your beliefs and vision can evoke a sense of shared group identity and potentially reinforce the sense that the group can effectively address environmental problems, thus helping construct a politicized group identity cf. But the proliferation of new media and communication technologies is changing the nature of collective action with many groups existing online and with little or no face-to-face interaction amongst members. On the one hand these new ways of conducting collective action can reach large international audiences; online campaigns are sometimes viewed by millions of people and result in swift responses on the part of business and decision-makers.
This raises the possibility that this type of collective action may result in a stronger sense of efficacy than for example more traditional protests , although it also raises the question of what group identity it would foster. The social identity approach seems well-placed to provide a framework for investigating and understanding these new forms of environmental collective action e. Some approaches to building support for addressing climate change and environmental problems are coming from grass root movements that bring together people in small groups to build a sense of efficacy to change behavior e.
Although these grassroots approaches may vary in format, at their core is the notion that being part of the group will empower people to make changes to their own lives and to potentially become role models for others who are not members of the immediate group. Research framed by the social identity approach, though, has had a tendency to focus on large scale categories and groups e. The social identity approach has spent less time examining small group mobilization, and this might help explain why the work on grass roots environmental movements has emerged largely independently of the social identity approach.
To our knowledge these insights have yet to be specifically applied and tested in the context of environmental mobilization. The social identity approach therefore provides a depth and breadth of theorizing about group processes that could offer insights to maximize the effectiveness of these groups for changing environmental attitudes and behavior.
It is clear that social identity is a powerful influence on attitudes, beliefs, and actions relating to climate change and the environment more broadly. Understanding the influence of social identity on environmental decisions and behavior also suggest strategies to promote a more environmentally sustainable world. These include: 1 using ingroup messengers, 2 forging a superordinate identity to reduce intergroup environmental conflict, 3 linking identity and pro-environmental outcomes, and 4 promoting pro-environmental ingroup norms.
Past research has gone some way to providing the empirical evidence to support these claims, however, there is still some way to go in testing these social identity-based solutions as well as providing social identity insights to address environmentally significant problems. Both authors collaborated to research and write this review article. The first draft of paper was written by the KF and MH edited and contributed sections to the paper. The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
National Center for Biotechnology Information , U. Journal List Front Psychol v. Front Psychol. Published online Feb Kelly S. Hornsey 2. Matthew J. Author information Article notes Copyright and License information Disclaimer. Reviewed by: Jessica M. Fielding, ua. This article was submitted to Personality and Social Psychology, a section of the journal Frontiers in Psychology.
Received Oct 2; Accepted Jan The use, distribution or reproduction in other forums is permitted, provided the original author s or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms. This article has been cited by other articles in PMC. Abstract Environmental challenges are often marked by an intergroup dimension. Keywords: social identity, intergroup, norms, climate change, pro-environmental attitudes, pro-environmental behavior.
Introduction The seriousness of environmental issues currently facing the world is increasing despite substantial research attention and the efforts of local, national and international environmental organizations. The Social Identity Approach The social identity approach incorporates two interrelated theories — social identity theory and self-categorization theory — which each seek to explain how individual attitudes, emotions, and behaviors are influenced by the group memberships to which we belong.
The Social Identity Approach and Environmental Attitudes and Behavior Identity and Assimilation to Ingroup Norms As outlined above, when social identity becomes salient, similarities amongst ingroup members and differences between ingroup and outgroup members are accentuated. The Influence of Intergroup Conflict As we noted previously, negative and competitive intergroup relations may arise when ingroup members perceive illegitimate status differences between their own group and other relevant outgroups Branscombe et al.
Integrating the Social Identity Approach with Other Relevant Theories A question raised by our analysis is whether the social identity approach adds to the understanding of environmental problems beyond other prominent theoretical frameworks. Social Identity Strategies to Encourage More Positive Environmental Outcomes If we accept that the social identity approach offers a helpful theoretical lens through which to examine environmental attitudes and behavior, then it should also be able to offer solutions to address environmental problems and conflicts.
Table 1 Social identity strategies to encourage more pro-environmental attitudes and behaviors. Social identity strategy Example study Use ingroup messengers Ingroup sources are influential because they are perceived to be more trustworthy and credible by ingroup members Schultz and Fielding Conflict over watershed restoration was transformed through forging a superordinate identity in a collaborative learning setting that allowed consensus to emerge and recommendations to be developed Link social identity and pro-environmental outcomes Identifying with a pro-environmental group will lead group members to conform to the pro-environmental attitudes and behavior of that group Van der Werff et al.
Nolan et al. British participants thought of themselves as more pro-environmental when compared to the U. Group leaders who used inclusive language influenced group members support for renewable energy. Open in a separate window. Use Ingroup Messengers When thinking about how to promote more positive pro-environmental outcomes, an important consideration is where the messages come from. Forging a Superordinate Identity to Reduce Intergroup Environmental Conflict Conflict between groups on environmental issues has the potential to impede progress on addressing these issues. Promoting Pro-environmental Ingroup Norms Following on from the point above, there is strong evidence that people are more likely to act in environmentally friendly ways when the norms of a behaviorally relevant ingroup—especially one that people identify highly with—are supportive of pro-environmental action.
Future Directions Throughout this paper, we have highlighted interesting and important questions that could be pursued by researchers within a social identity framework. Conclusion It is clear that social identity is a powerful influence on attitudes, beliefs, and actions relating to climate change and the environment more broadly. Author Contributions Both authors collaborated to research and write this review article.
- Philosophy of Economics (Stanford Encyclopedia of Philosophy)?
- Hells Harbor (Action Adventure).
- Shop by category.
Conflict of Interest Statement The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest. References Abrams D. Knowing what to think by knowing who you are: self-categorization and the nature of norm formation, conformity and group polarization.
The theory of planned behavior. Efficacy of the theory of planned behaviour: a meta-analytic review. Promoting pro-environmental action in climate change deniers. Social identification, social representations, and consumer innovativeness in an organic food context: a cross-national comparison. Food Qual. The ironic impact of activists: negative stereotypes reduce social change influence. Aspiring to mitigate climate change: superordinate identity in global climate negotiations. Public division about climate change rooted in conflicting socio-political identities.
Oxford: Blackwell Science;. The impact of norms and assumed consequences on recycling behavior.