How Power Relationships are Involved in Research Methods

With the rise of activism and activist research, this paper explores how power relationships are involved in traditional and emerging methods used in research on activism. This question matters as research methods have the potential to both improve the capacities of activist groups and enhance knowledge of agents involved: researcher and activist. The added value of the paper is that it presents a range of methods used in research on activism, including new methods that are relatively uncommon in planning research. The second contribution of this paper is that it is based on a power framework by Forester; it analyses how power is embedded in the use of a particular research method. The authors find extant differences between the methodologies when analyzed through this framework, especially in their potential to involve with activist communities. The authors encourage researchers to be braver in using activist research methods and to be aware of the underlying power discourses in their choices.


AESOP / YOUNG ACADEMICS NETWORK Introduction
Within cities and public spaces, activism is increasing. Examples of this increase include the emergence of tactical urbanism (Mould, 2014), austerity urbanism (Peck, 2012), social movements (Schoene, 2017) and insurgent spaces (Roy, 2005). Activists challenge the tacit, neo-liberalist assumptions behind many plans, spatial interventions, regulations and policies. Subsequently, they challenge the monopoly of urban planners in triggering socio-spatial transformations (Sager, 2016). In this way, activism challenges and changes existing power relations, social norms and values.
Therefore, it is no surprise that this rise in activism has drawn the interest of several planning scholars (e.g. Sager, 2016;Mould, 2014;Scholl, 2017, p. 46). A central characteristic of planning research on activism is that such research aims to benefit the powerless and expose (unequal) power relations between urban planners and activists (Cancian, 1993). It may provide a new perspective to researchers about how local communities and groups of activists can be included in academic research. The involvement of participants in research projects may improve understanding of activism as a transformative practice of communities and spaces, as well as develop the (research) skills of activists.
While urban planning researchers have lately started to study activism, little is known about the methods researchers use. This is problematic, as it can lead to the use of methods 1) that may be less useful in understanding activism and 2) that draw information from activists for answering the research questions of researchers, while not boosting the knowledge or skills from activists. As answering research questions is only possible when the correct method is chosen (Baarda et al., 2013), it is important to understand the various methods planners could use and their advantages and disadvantages. Otherwise, it may result in a lack of insight into activism as an important force of spatial and social transformations. For instance, tactical urbanism initiatives are typically local, short-term implemented, resource-scarce changes of the streets and plazas that make up a city. They are interesting for researchers as they challenge formal, bureaucracy-led planning and involve small scale changes of public spaces that benefit locals (Silva, 2016). To fill this gap, more attention needs to be paid to the full range of research methods that exist: methods which can be used to study activism and the knowledge they produce for both the researcher and the activist. This paper fills this gap.
Planners studying activism frequently opt for more traditional scientific research methods, such as interviews or observations. However, a range of potential activist research methods has been developed that might be useful in expanding the methodological repertoire of planners. The focus will concentrate upon six research methods that are used by planners to study activism: participant observations (PO), (semi-structured) interviews, surveys, Community-Based Participatory Action Research (CBPR), Participatory Action Research (PAR) and Virtual Reality (VR).
The authors explore the extent to which power relations play a role within these methods. Power plays an important role in carrying out activist research, as conducting research involves a reciprocal relationship between the researcher (the user of a method) and the activist. Furthermore, research methods can be used to expose unequal power relations. However, as will be discussed, some methods are more able to do so. Finally, methods can be used by researchers to enhance the power of activists. To analyze the role of power existing between the planning researcher and activist, the authors apply a well-known framework by Forester (1988). Finally, the paper indicates that more traditional methods do not enhance the capacity of activists, whereas these new research methods are more able to do so. Furthermore, the authors plea that planners might be bolder in selecting their methods.

AESOP / YOUNG ACADEMICS NETWORK
The paper follows a structure which begins by outlining the six research methods that the authors identify as most familiar to the study of activism; in planning; the second section explores Forester's framework of power. Next, the authors link Forester's framework to these six research methods to explore how these empower activists.

'Classic and Frequently Used' and Activist Research Methods
The authors interpret 'classic and frequently used' methods here as; participant observations, surveys and (semi-structured) interviews and as such it is a categorization through which to compare. These research methods continue to generate robust results and are useful. Therefore, the authors do not consider these methods to be outmoded. However, there is a range of new methods developed in other research areas (e.g. public health: CBPR) that might complement observations, surveys and interviews in conducting activist research.
The first research method is observation research. Participant observations (PO) may involve a researcher observing the activities of activists, without actively intervening or controlling in the observed situation. Such observations are conducted by a form or protocol and logged afterwards (DeWalt & DeWalt, 2011, p. 1). An important advantage of POs is that they allow researchers to study activism in a socio-spatial context: in the street, neighborhood, park or city where it therefore occurs and affects the surrounding community. However, the researcher does not intervene and quietly observes what occurs. A disadvantage is that they only describe what participants are doing, not why they do something. Another deficiency is that some behaviors are relatively rare; such as spontaneous criminal activities and certain other forms of activism (e.g. graffiti) (Baarda et al., 2013). Thus, many forms of activism may be missed and undocumented.
However, what is highlighted in particular is the process and relationship between the researcher and participants. Planners as researchers participate in activities or events to observe events and subjects. As a result, they operate from an external point of reference, necessary to enable them to observe situations and processes but not to influence them. This makes the scientist a ´silent observer´. A potential difficulty is that the researcher might misinterpret the behaviors of activists, which may lead to false conclusions (Allmendinger, 2008).
The second method used to study activism are surveys (e.g. Knigge, 2009;Scardaville, 2005). As a frequently used method, a questionnaire is used to generate information about a larger population by a smaller sample (Kaase, 1999, p. 11). The advantage of surveys is that a lot of data can be collected in a relatively short time and is less likely to trigger socially-desired responses (as surveys are generally anonymous). Furthermore, surveys offer the possibility to establish causal links and make inferences about the broader population. However, surveys can present disadvantages. One weakness may be a difficulty in reaching particular target groups (e.g. activists) or when participants of the survey are not representative of the broader population (i.e. sampling bias). Furthermore, the format and quality of surveys are highly dependent on how the questions are presented and understood and lastly, only a limited amount of questions can be asked (Baarda et al., 2013).
A third and common method in social science research are qualitative, semi-structured interviews (QIs). The interviews are often conducted by pre-established interview guidelines (Alsaawi, 2014, p. 151). Semi-structured interviews allow researchers the opportunity to gain in-depth knowledge and richness in responses (Bryman, 2008in Alsaawi, 2014) of experiences, beliefs and norms of activists. Their open nature retains the potential for activists to bring their topic to the table and in this way create innovative knowledge that is unfamiliar

AESOP / YOUNG ACADEMICS NETWORK
to the researcher. On the other hand, this method only succeeds if there is a degree of 'rapport' between the researcher and activist, i.e. the degree of harmony and trust in a conversation (Hennink et al., 2010;King & Horrocks, 2010, p. 3). Furthermore, semi-structured interviews reveal the truth of the activist, but this is a subjective account of what occurred. Moreover, causal links cannot be made using this method. Finally, in the case of sensitive topics, it can be difficult to recruit respondents. Although each of these three methods is frequently used to study activism (e.g. Ginwright & Cammarota, 2007;Knigge, 2009), they represent only a few on offer. In addition to these established methods, three emerging research methods in the field of activist planning research are examined: PAR, CBPR and methods based on VR.
PAR is an approach in which action and social change are central. PAR stresses a strong collaboration between participants: researchers and activists (Burns et al., 2011, p. 15). PAR involves a reciprocal exchange of knowledge, skills and power: researchers enhance activist's knowledge of research skills, while activists provide knowledge and resources about the community (Kim-Ju et al., 2008). One way of doing this is by creating a joint project, such as using music, painting or theatre to bring activists, researchers and others together (van der Vaart, 2018). The use of music, painting or theatre has the potential to create interactions between the creators, users and planners and within communities themselves (Simonsen et al., 2014). Participants are given space for critical self-reflection and analysis of reality to generate more authentic knowledge (Gaventa & Cornwall, 2015, p. 470). One disadvantage of PAR is the time and effort required of both the researcher and activist. Another disadvantage is its focus on individuals (activists) that are part of the community, not the community itself as the object of study (Kim-Ju et al., 2008). This focus on the individuals in a community is a key difference with the next method CBPR.
CBPR is a research method directly focusing on the relations between academic and activists (Wallerstein & Duran, 2006, p. 1). According to Minkler & Wallerstein (2003), academic and activist partners are involved in the research process: they collaborate in the creation of the research (sub)question, as well as the data collection and analysis. Thus, the research questions and process is the responsibility of both the activists and researchers: the researcher somehow becomes a part of the activist community. This method seeks to change the roles of the researcher and stakeholders (Burns et al., 2011, p. 5). CBPR is a place-based research method, which focuses on an activist community, not the individuals. With a combination of different data, for example, from interviews, focus groups or mappingprocesses, again conducted by both researchers and activists, the researcher receives their results (Burns et al., 2011, p. 6).
Another activist technological-method is VR, which can be used to visualize planning processes or scenarios, for example, for activists to illustrate future developments (Portman et al., 2015). With the help of future artificial spaces, activists can actively experience possible effects of their actions. For example, computer simulations can imitate experiments and make them ´real´ (Portman et al., 2015;Natapov et al., 2016;Psotka, 1995). 3D green spaces, as well as virtual square or park designs, can be experienced. This new methodology is beginning to be utilized (Psotka, 1995, p. 405). With the help of this technique, participants can explore and engage with their environment in a new way. It can provide both visual but also other sensory stimuli and offers a range of data and presentation for both the activists as well as the researcher. However, VR is continually being upgraded and rapidly evolving, and so the uncertainties in these processes must be clearly articulated (Portman et al., 2015, p. 381). A disadvantage of VR can be that it often takes place in laboratories, and therefore the behaviour of people in real space cannot be understood. One advantage is, however, that virtual developments can be depicted realistically.

Forester as a lens to understand power
The authors use Forester (1988, p. 144) to map how power relations are intertwined while conducting activist research (see Table 1). Forester perceives power as a sociallyconstructed, reciprocal relationship between two or more agents that is reproduced and changed by human (inter)action (Tait & Campbell, 2000). Forester provides four criteria to analyze power relationships. While Forester's criteria are more broadly applicable as they involve the power relations between agents, the authors translate these criteria to operate at the level of the power relationship existing between the planner-researcher and the activist (community). In operationalizing, the authors followed two rules: 1) the interaction was between a researcher and activist, 2) the researcher selects and executes a particular method, and 3) the researcher is influenced by the standards and ethical codes of conduct in academia. Truth Information given to activists is factually correct. Information given to the researcher is correct. Information is given to correct misunderstandings or improve the knowledge (and consequently their agency) of activists.

Legitimacy
Legitimacy of the method used.
Taking advantage of having more knowledge. Ethical norms in research.
Forester distinguishes four key criteria to analyze power relations: comprehensibility, sincerity, legitimacy and truth. As comprehensibility, Forester means that activists should be able to understand the questions they are asked to them and the purpose of the research. It involves how questions are formulated, and how the researcher presents themselves. Trust is about the degree of trust between researcher and activists. It entails the degree to which researchers are neutral observers or favor particular world views (e.g. the researcher as an advocate of neoliberalism). Legitimacy refers to the extent activists perceive the goal of the research and the use of particular research methods as being legitimate. It involves the ethical rules and norms that come with using a method, such as not misleading participants (activists) or taking advantage of having more knowledge. Finally, truth is about the degree to which the claims, beliefs and information given by the researcher and activist correspond with the factual truth. This criterion involves the degree to which activists gain (accurate, factually correct) information from the researcher. It is important to acknowledge that Forester (1988) stresses that these criteria can never be fully satisfied.
Balancing the elements of the criteria play an important role in researching activism. Firstly, the criteria help illustrate how the use of a research method is perceived and experienced by the activists. Secondly, the criteria can help in determining how the selection and use of a research method is an act of exercising power. Besides, the researcher is bounded by the ethical and moral academic codes, constraining their conduct and selection of methods during

AESOP / YOUNG ACADEMICS NETWORK
the research (Allmendinger, 2008). Finally, it helps explaining how the use of different methods can enhance the agency of activists, as they might learn from the research process and outcomes of the research processes.

Analyzing Methods: How are They Linked to Power?
In PO, the researcher does not have any direct involvement in the activist community. The researcher retains an outsider perspective and does not -in any way -limit, direct or influence the actions of activists. If the method is well applied, the purpose is clear. This makes the method comprehensible for the activists (comprehensibility). If the participants feel unobserved in this method, they express themselves without any constraints (Baarda & De Goede, 2013, p. 250). This may be problematic when the observed behaviour concerns illegal actions and may lead to the incrimination of activists. They express their opinions openly and so, the reality of the activist can be revealed (sincerity & truth). The methods can be used in certain research contexts only and may not be suitable for all topics due to the passive role of the researcher and the rareness of certain behaviors/actions (legitimacy).
The second method the authors consider is surveys. Concerning comprehensibility, surveys can be problematic. The formulation of the survey questions can unconsciously steer participants towards certain answers or lead to misunderstanding (e.g. Rooney et al., 2005). As it is generally not possible for participants to ask for clarifications, researchers must be cautious in how they formulate questions. Pilot testing can help to improve comprehensibility for activists (Baarda & De Goede, 2008). Subsequently, sincerity might also be problematic. Researcher and activist are not equal in a survey: surveys are a rather one-way method that draws data from activists, while activists generally do not have much influence on the questions asked. Also, the framing of these questions can contain a hidden agenda or steer participants towards certain outcomes. Moreover, the legitimacy of surveys is dependent on the research topic. Surveys can be less legitimate for sensitive research topics, as less trust can be built between the researcher and the respondent than in other methods (legitimacy). This may mean that participants will choose not to reveal personal details about sensitive topics. Finally, the truth criterion is a strong advantage of surveys. Surveys can be used for both qualitative and quantitative research questions and subsequently provide the opportunity to answer a wide range of questions (Jansen, 2010). An important factor further contributing to 'truth' is that surveys offer the possibility to conduct statistical analysis (e.g. in SPSS) to improve their accuracy and establish causal links (Baarda et al., 2014).
In interviews, the interviewer and the activist experience a face-to-face connection. For this reason, they both can enter into a more 'relaxed' form of a conversation devoid of formal language and challenging questions. The interviewer has the chance to clarify situations, themes and topics for the interviewee to prevent any misunderstanding (comprehensibility). If the relationship between the interviewee and the interviewer is built on trust, interviews permit the researcher to gain contextual meaningful information about how the real world is perceived (Denzin, 2001in Alsaawi, 2014) and provide in-depth and creative knowledge (Shallwani & Mohammed, 2007, p. 31) about the activist. Both the researcher and the activist are equal partners in the interview (sincerity & truth). Interviews impress with their flexible and open design (Alsaawi, 2014, p. 154) and can, therefore, be adapted easily. On the one hand, the researcher has the power regarding his research (ibid.) and thus must act ethically while doing it (legitimacy). On the other hand, ´the interviewee has power as a "privileged knower"´ (Nunkoosing, 2005, p. 699 in Alsaawi, 2014).
The first activist method is PAR. PAR can be considered as being comprehensive. The main argument is that the research questions and data collection occur in cooperation with the

AESOP / YOUNG ACADEMICS NETWORK
activists. The involvement of activists may mean that the research questions and methods are selected and equate with those of the community which is the subject of research (Kim-Ju et al., 2008). The use of music, painting or theatre can also make PAR understandable for a large number of actors (comprehensibility)(van der Vaart, 2018). However, a potential danger for researchers of PAR may be that they lose their neutrality, as the researcher may become too involved with the activists. This may be problematic, as it may lower the criticality of the researcher towards the activist community (Levinson, 2017). However, a potential danger of PAR is the exclusion of certain elements or activists, as some activists may not be willing to participate due to time or financial constraints (Levinson, 2017). This method is especially legitimate for research topics in which trust of the activists is necessary, as this method starts with the building of trust between researcher and activists (legitimacy). Finally, trust is necessary, especially for sensitive research topics, such as gaining an understanding of illegal activities (Ochocka et al., 2010). When trust exists, this method may teach activists new research skills and knowledge, while researchers learn about the way activism is carried out whereby the method can be legitimized (legitimacy; Kim-Ju et al., 2008). As with all qualitative methods, PAR can reveal the perceptions, norms and beliefs of participants, but not the factual truth or establish causal links (Hennink et al., 2010). Whether the revealed perceptions and norms are 'true', dependent on the degree of trust between researcher and activists, is crucial.
CBPR is based on the assumption that complex problems cannot be solved by experts alone. Researchers need key insights provided by a group (Burns et al., 2011, p. 5). An advantage of CBPR is that the knowledge is produced at a low-threshold level. The actors work together on decision-making processes on an equal basis. Since all partners should have, at least in theory, equal rights, they can share their interests and fight for them in the research process. If the process is well-organized, this method results in a broad scope of action for all actors (comprehensibility), which makes it possible to participate honestly and improves transparency (sincerity). Furthermore, by cooperating with researchers, activists may improve their understanding and knowledge, for instance about how research questions are formulated or a research method is used. This stimulates activists' agency (truth). Also, most spatial interventions that come from outside do not often create the desired results (Burns et al., 2011, p. 5) because of a lack of understanding of the case specifics. Moreover, on the one hand, a challenge is to create and maintain legitimacy between the participants, as tensions or conflicts may arise during the research process (between researchers and activists). Especially as CBPR focuses on the long term, such conflicts or differences in opinion may emerge. On the other hand: ´there is value and legitimacy in knowledge of individuals, families, and others in the community´ (legitimacy, Burns et al., 2011, p. 5).
VR methods make it possible, primarily through various visualizations and new techniques of representation and experience, to put people into situations more easily and sometimes even recognize the immediate consequences of their actions (Portman et al., 2015;Natapov et al., 2016,;Psotka, 1995). Thus, the threshold of participation is relatively low and makes the process understandable for various people (comprehensibility). The challenge, however, is how the potential results can be linked and generalized to the non-virtual world or activist community. One difficulty for certain groups and the researcher can be to generalize the results because they have to legitimize the goal of the research and stick to ethical norms (legitimacy). This method involves a high degree of uncertainty in the process, because it is based on a virtual world, which must be clearly articulated (Portman et al., 2015). Otherwise, participants are not be able to use the method adequately or understand it, as building a virtual or 3D model requires much technical knowledge and skills. VR-based methods do not enhance the knowledge or capacities of activists (truth). With regard to criteria four, this aspect can overtax activists. This could limit their room for action because they may not feel as equal partners in the research process due to the lack of knowledge about virtual technology (sincerity).

NEXT GENERATION PLANNING
Open Access Journal 28 AESOP / YOUNG ACADEMICS NETWORK Face-to-face interaction without formal language make the methods comprehensive.
This method can result in a broad scope of action for all actors.
The involvement of activists in the research process may improve the comprehensibility for them.
People can put themselves into situations more easily and sometimes even recognize the immediate consequences of their actions (virtually).

sincerity
Participants conduct their behaviour openly, without any perceived constraints (as the researcher is a silent observer).
Surveys offer the possibility to analyze the results of surveys statistically (e.g. SPSS) to improve their accuracy and establish causal links.
Interviews permit the researcher to get meaningful contextual information to the real world. Researcher and activist are equal partners.
If the process is wellorganized, participants can work honestly. This helps to be transparent in the (research) process.
If the actors can not use the method sufficiently or even understand it, they feel overwhelmed. Using this method may require many technological skills. This aspect can also overtax them and limit their room for action.

truth
Method starts with the building of trust between researcher and activists.

legitimacy
The methods can be used in certain research contexts only and may not be suitable for all topics due to the passive role of the researcher and the rareness of certain behaviors/actions.
Surveys are less useful when the research is about sensitive research topic (requiring high degree of trust between researcher and activist).
The researcher has the power regarding his research and therefore should act ethically while doing it.
On the one hand, a significant challenge is to create and maintain legitimacy between the participants. But on the other hand, ´there is value and legitimacy in the knowledge in the community.
The method may teach activists new research skills and knowledge, while researchers learn about the way activism is carried out.
The challenge is to link the results with the non-virtual world and to legitimize the goal of the research.

AESOP / YOUNG ACADEMICS NETWORK
While each method has its advantages and disadvantages, the analysis and framework of power (see Table 2) shows that the activist research methods are more able to enhance the agency and capacities of activists, compared to interviews, surveys and PO. However, these activist research methods may run in opposition to academic standards and conventions (Cancian, 1993, p. 92;van der Vaart, 2018). Therefore, the power of these research methods for academia may be lower, as using them may make it hard to publish in academic journals (Mark Chesler in Cancian, 1993, p. 105). However, it presents the opportunity of greater community involvement and more societal impact, which is in the end what planning is about (Cancian, 1993, p. 105).

Conclusion
In summary, the paper aims to illustrate recent and innovative methodological choices against those which are more established, more routinely utilized and those emerging in recent research. When presented together they reveal differences in underlying power discourses which are embodied in the various methodologies. And when examined through the lens of Foresters' (1988) criteria of power, it appears that the dimensions of power operate unevenly throughout the selection of methods. Forester distinguishes four key criteria to analyze power relations: comprehensibility, sincerity, legitimacy and truth.
The authors demonstrate that the first criterion highlighted by Forester, comprehensibility; involving the degree to which the participants understand the questions asked of them and accordingly the purpose of the research rests with the skill of the researcher in the case of both classic and frequently used and newer methods. However, due to greater involvement of the activist and co-creation of methods from the beginning of the research for activist research methodologies, the authors would argue that one can see a higher trend of comprehensibility in the activist research methods. The same may be argued for the criterion of sincerity, as equal partners; researcher and activist and attainment of neutrality when carrying out the research showed more strongly for innovative research methods due to the nature of participation and ability of activists to respond honestly. This, in turn, creates a level of transparency between researcher and activist. For the criterion of legitimacy, where the authors found that use for future research processes depends on the context in which it is intended for dissemination. While in the case of the newer methods, legitimacy is established and a value placed if the knowledge is returned into the community. However, questions remain around the suitability of methods for publications within academia, which may contradict the efficacy of the research and threaten legitimacy. In so far as qualitative methods deal with perceptions, norms and beliefs of participants but lack an absolute truth, the criterion of truth is maintained through an agreement of shared factual truths and the correction of misunderstandings. Neither surveys, PO's nor VR's, offer this possibility, whereas CBPR, PAR and interviews can.
The authors have highlighted that there are opportunities but also risks associated with the relations of power within the various methodologies. For researchers, this choice should be investigated and evaluated in light of these dimensions of power to help determine their selection. The authors hope to have illustrated the need to develop awareness across a spectrum of characteristics existing within the choices of methodologies but also to encourage greater insight into the practice of activism which can be achieved through a careful choice of methods. Ultimately, the researcher can enable activist efforts which place the researcher in a more active position, not only as a passive bystander but critical to wielding power equitably. Conclusively, through a careful selection of methodologies, 'brave' and innovative choices are possible.