The Victim’s Right to Intervene as an Injured Party in Criminal Proceedings A Multidimensional and Interdisciplinary Assessment of Current Dutch Legal Practice

In order to further crime victims’ compensation, the Dutch legislator relatively recently extended the admissibility criterion. Since 2010, the key lies in the assessment of whether such a claim presents an undue burden for the criminal proceedings (S. 51f DCCP). In order to learn whether this legal change has been effective, an evaluative research was carried out containing both legal theoretical and empirical research (quantitative and qualitative). This called for an accurate research design in which classic legal research methods need to be combined with methods applied in social sciences. To counter this methodological challenge, we opted for a mix of research-design methods, using triangulation as the key to enable us to analyze the ‘law in action’. Our approach turned out to be fruitful, although not without methodological hazards. The paper contains a report on our ‘methodological journey’, providing an account of the difficulties that we encountered during the execution of our mixed research methods. The aim of the paper is only modest: we want to report on ‘the lessons learned’, endorsing the fact that triangulation is an appropriate approach for an interdisciplinary assessment of Dutch legal practice.


Introduction
This contribution provides a description and an evaluation of the methodological design of a recent study of Dutch legal practice with regard to the victim's right to intervene as an injured party in criminal proceedings. The aim of the article is to gain a further insight into the opportunities and difficulties of conducting empirical-based legal research using a mixed methods approach. 2 Given the current tendency within legal research to combine classic legal research methods with methods applied in social sciences, there is a great need to reflect on how to address the issues that we face in this regard and to search for methodological approaches to combine the best of both worlds. 3 The recent research that we have undertaken on the issue of civil compensation in criminal proceedings will be used to illustrate the approach which we have used to build an accurate research design in which classic legal research methods are combined with methods that are applied in social sciences, the added value of studying legal practice using (qualitative and quantitative) methods that are applied by other disciplines and the difficulties we encountered in doing so. 4 The paper is structured as follows. First, we will further elaborate on the current tendency to incorporate insights from other disciplines into legal studies and explain what problems have been recently addressed with regard to this exercise in academic literature (Section 2). Next, we will introduce triangulation, which in our opinion is a useful approach to bridge the gap between the different disciplines as it may well help mono-disciplinary legal research design will represent a lesser methodological challenge. 14 However, things can become (more) complicated if one combines multidimensional and interdisciplinary approaches. 15 The former implies a cross-border analysis from a commonly shared internal perspective, e.g. the normative perspective of 'the law'. Combining such an approach with an interdisciplinary one obviously adds an extra layer to the methodological issues that need to be resolved. Taekema and Van Klink describe the latter as 'aiming at an integration of disciplines'. 16 This means, according to Taekema and Van Klink, that 'the central question, the methods, and conclusions are jointly developed and motivated by both disciplines'. 17 Both avenues, however, provide fruitful ways to explore the phenomenon that one aims to study. Moreover, combining a multidimensional and an interdisciplinary approach implies a more holistic study, evoking contemplation over underlying (scientific) narratives. 18 Legal research, being traditionally a normative discipline, is in need of such contemplation and reflection. Contemporary social changes entail major challenges for the legal authorities in how to legitimately maintain social order, indicating a need to study the law in theory and practice. This implies a need for legal researchers to open out towards interdisciplinary research, next to executing the multidimensional, comparative perspective that they developed during past decades. Notwithstanding the strong nature of his wording, Bodig's statement that 'interdisciplinary scholarship offers pretty much the only feasible way to preserve the competence of legal scholarship in exercising some control over doctrinal development' does contain a grain of truth in light of the contemporary instrumental, policy-related use of the law. 19 As a result, legal researchers are nowadays confronted with results from other empirical-based studies that are somehow linked to their central research question. This does not imply that fundamental theoretical legal research has lost its value. 20 On the contrary, legal theoretical research lies at the heart of legal methodology, both disciplinary and mono-disciplinary. 21 It is precisely because of the constant flow of extra-legal information causing doctrinal knowledge to be in a permanent state of renewal that the incorporation of empirical-based studies within legal research should be seen in the light of the normative nature of the questions addressed. 22 Taekema's classification of legal research as a 'relativist autonomous' discourse reflects this dominant 'pragmatic perspective'. 23 To phrase it differently: legal research aims to study social developments and to provide for a legal interpretation in order to enable legal innovation.
Nevertheless, we need to be careful not to turn a blind eye to the caveats that are present within such a perspective. 24 What we may achieve is a more coherent image of 'reality' being created, combining data that relate to distinct aspects and layers of the social phenomenon addressed. 25 If we choose to opt for such a multidimensional research design, we need to reflect on the nature of our competences. 26 Here Giesen raises the question of 'whether it is in fact possible -and if so, how, why and when -to leap from such extra-14 See Fielding (2012), supra note 2, p. 125. Also Taekema & Van Klink, supra note 3, p. 2560. 15 'Multidimensional' relates to starting from an internal, in our case, legal perspective and is distinct from 'interdisciplinary' indicating cooperation -to a certain extent -with social science (see Taekema This question refers to the stage during  which a legal researcher, who has studied both traditional sources of legal research and the results of studies  from other disciplines, has to answer the normative central research question. 28 Here Giesen's questions  come to the fore: what role do the results, taken from the other disciplines, play at this stage? And how does one build normative legal conclusions on the basis of multidisciplinary 29 or interdisciplinary research? 30 In the research which we have undertaken, we were not only faced with this last-mentioned problem of how to build a bridge between non-legal insights and answering a normative, yet practice-related legal question. Our study, as we will further explain in Section 5, aimed to evaluate the current legal practice with regard to victims' compensation. The request by the Research and Documentation Centre, which commissioned the research, included a suggestion to conduct interviews and/or a file research. This gave rise to the methodological question of how to combine these methods that are not particularly applied in legal research with traditional legal methods. In this regard the socio-scientific approach of triangulation helped to design an adequate research plan to answer the evaluative central question. We will now clarify the approach of triangulation as used within this research and then we will introduce the subject-matter of the research and its plan (Sections 4 and 5).

Triangulation
Triangulation is not an undisputed concept, especially in light of the contemporary mixed methods debate. Fielding mentions the 'triangulation metaphor', and points out that triangulation is a 'practical' approach, rather than a technique. 31 As is the case in every research, attention needs to be paid to the underlying epistemological precepts underpinning the method(s) used. Triangulation may not be used as a means to cover up methodological deficiencies. Nevertheless, if it is properly executed, triangulation provides 'a theory of variance' that enables legal researchers to understand legal phenomena by assessing variables and correlations in terms of events and processes. 32 In the social sciences triangulation is described as the mixing of data, methods or researchers so that diverse perspectives cast light upon a research topic. 33 The idea of triangulation originates from quantitative research (e.g. Campbell & Fiske in 1959) 34 and was made popular in qualitative research (Denzin in 1970). 35 In Jakob's words, 'by combining multiple observers, theories, methods, and empirical materials, researchers can hope to overcome the weakness or intrinsic biases and the problems that come from single-method, single-observer, single-theory studies. Often the purpose of triangulation in specific contexts is to obtain confirmation of findings through convergence of different perspectives. The point at which the perspectives converge is seen to represent reality.' 36 Moreover, combining different data sources and both qualitative and quantitative research would contribute to validity and reduce the effect of the peculiar biases of each research. 37 Data triangulation, in other words combining data types, would increase the validation of the outcomes. Attempts to gather data through different sampling strategies ensures that a theory is tested in more than one way, for example combining data from interviews with observations and document analyses contributes to internal validity. This is closely related to methodological triangulation in which two or more research methods of data collection are combined within a single study. This usually involves a combination of qualitative and quantitative research. 38 According to Webb et al.: 'When a hypothesis can survive the confrontation of a series of complementary methods of testing, it contains a degree of validity unattainable by one tested within the more constricted framework of a single method.' 39 The latter implies researchers' triangulation. Next to the need to prevent preconceptions and to further reflectivity ('two heads are better than one'), triangulation mostly implies co-operation amongst researchers from various disciplines. 40 Applying empirical research with regard to a legal phenomenon therefore requires the presence of researchers from different scientific backgrounds. The latter is especially true in the case of adding a quantitative aspect to the research design since legal researchers, as a rule, are not trained to use such techniques. To a lesser extent, the same holds true for the qualitative part of legal research, at least the empirical part thereof. Having been trained in hermeneutics, conducting interviews is apparently not a major challenge for legal researchers given the shared 'linguistic' element. Nevertheless, serious methodological issues may occur since traditional legal education does not provide for interview techniques or how to analyse such qualitative data. Researchers' triangulation then prevents misinterpretation. However, this comes at a price since triangulation requires the researchers involved to develop a communal understanding of the concepts used, especially the legal concepts that are often in need of constant interpretation Triangulation thus places a heavy burden on inter-collegial co-operation. Nevertheless, it also provides opportunities to clarify one's (legal) preconceptions and to reflect on any potential bias.
Thus, from this perspective, combining different data sources and both qualitative and quantitative research may have an important value in studying legal topics. Indeed, triangulation has been criticised due to its rather open nature. 41 However, while triangulation seems to fit well in research methods with common ontologies and epistemologies, if different assumptions are used, problems may then emerge. 42 What does one do if the descriptions of the realities contradict each other? From a social constructionist approach, there is not one truth; it is rather the case that different perspectives provide insights into different meanings and realities. The social construction of the law illustrates this restriction. 43 Indeed, the current profile of victims as rightful claimants of statutory rights within criminal proceedings represents such a social construction. Had it not been for the presence of social(-economic) developments, giving way to individualization, authority crises and the rise of interest groups (especially the women's movement), the victim would not have become the focal point of Dutch criminal politics. 44 To conclude, Figure 1 may serve to visualize our approach, applying data and methodological triangulation in our study to cover the relevant aspects of the previous issues addressed. 38 M.E. Duffy, 'Methodological triangulation: a vehicle for merging quantitative and qualitative research methods', (1987)

Compensation for crime victims through criminal proceedings 45
Leaving aside triangulation as such for a moment, the time has now come to introduce the subject of our study, as we will hereinafter further explain the added value of triangulation from what we have learned while conducting this research. The subject of our study is victims' compensation. In order to further our comprehension, a brief elaboration is provided with regard to contemporary Dutch practice.
Research has shown that victims very much appreciate compensation in order to overcome the aftermath of crime. 46 Compensation for victims being a focal point of the current Dutch criminal justice policy, the authorities have established arrangements to promote compensation in the context of criminal proceedings. Based on the principles of fairness and procedural efficiency, the Dutch Code of Criminal Procedure (Wetboek van Strafvordering/Sv; hereinafter DCCP) traditionally provides the right to intervene and claim compensation (the so-called adhesion procedure (voegingsprocedure)).
The right to intervene is laid down in Article 51f DCCP. Although it is an element that is incorporated in criminal law, the assessment of the claim is done on the basis of the private law of obligations (in particular Section 6:162 Dutch Civil Code (Burgerlijk Wetboek/BW; hereinafter DCC)). The victim's right to intervene is thus in a way a mixed legal procedure. This mixed character is reinforced by the fact that the judiciary -who are to assess these civil claims -are predominantly criminal lawyers.
Because of the public-law character of criminal proceedings, the civil claim is ancillary to the assessment of the crime. In line with this, the admissibility of the victim's claim depends on an evaluation as to whether or not the ruling represents an undue burden (onevenredige belasting; Section 361 paragraph 3 DCCP) for the criminal proceedings. This criterion, introduced in 2011, is, however, rather vague. Although the Explanatory Memorandum provides for some clarification, the courts are left to provide a further interpretation.
From the alteration of the admissibility criterion in 2011 onwards, victim compensation has remained a focal point in Dutch criminal justice policy. In addition to the widening of the admissibility criterion with the intention of allowing more claims, policy measures were taken to promote assistance for crime victims to enable them to submit their claims through the Victim Support Desks (Slachtofferloketten), a joint effort by the police, the Public Prosecution Service (Openbaar Ministerie) and the Victim Support the Netherlands

Legal document analyses
Legal opinions

Qualitative data
Triangulation

Quantitative data analyses
Quantitative data (Slachtofferhulp Nederland). In addition, victims of serious crime were provided with a right to free legal aid. Already in the mid-1990s, the compensation order (schadevergoedingsmaatregel; Section 36f Dutch Penal Code (Wetboek van Strafrecht), 1995) was introduced, serving to relieve the victims from the burden of having the verdict executed. In addition, the Advance Payment Scheme (Voorschotregeling; 2006) was introduced and it provides that victims of violent and sex crimes, under certain circumstances, can receive an advance payment of damages by the State. In 2011 the scope of this scheme was extended to other crimes. Both measures represent major practical advantages for crime victims in pursuit of compensation. Other advantages for victims who intervene in criminal proceedings are that they do not have to pay a court fee and may rely on evidence that has been collected by the Public Prosecution Service.
The aforementioned alterations relating to the adhesion procedure do not stand alone, but are part of a broader policy to extend the victim's position within the Dutch criminal justice system. Over the past decade, the Dutch have rigorously extended the victim's position by acknowledging that the victim is a participant within the criminal proceedings and by granting the victim statutory rights. 47 The dilemmas that go hand in hand with the extended victim position resonate within the decision-making with regard to victims' compensation.
Given the major executorial advantages that accompany victims' compensation within criminal proceedings, it comes as no surprise that substantial use is made of the adhesion procedure on an annual basis. Previous research has shown that crime victims have a clear preference for compensation through criminal procedure, rather than lodging a claim with the civil courts. 48 Annually, around 18,000 claims are lodged, representing about a third of the total annual number of criminal cases. These numbers indicate that the Dutch judiciary and the Prosecution Service are regularly called upon to assess claims for wrongful acts in criminal procedures. With the latest evaluation of the adhesion procedure dating from 2007 49 and given the policy measures and legislative changes since then, the question raised is how victims' compensation is currently organised as part of criminal proceedings in terms of judicial decision-making and what the outcomes of the proceedings are. These were the key questions in the research design that will be outlined in the next section.

Central question
The central research question in our research was: In what way is victim compensation currently the outcome? 50 Although not explicitly mentioned in the central research question, it was clear from the beginning that an important aspect of the research was to discover the effect of the aforementioned alteration of the admissibility criterion in 2011. Moreover, it was clear from the subquestions that the central question implied that a comparison had to be drawn between today's legal practice and the status quo that was described in a similar report in 2007. 51

Interdisciplinary approach
The subject of our research, victims' compensation, evokes strong community opinions, and normative expectations lie ahead. To draw a picture on the basis of case law, parliamentary dossiers and legal doctrine would not do justice to the (social) problems involved. In order to be able to check, back up and explain 47 Amongst others, the right to make a victim's statement of opinion (S. 51e DCCP the 'theoretical' results, an interdisciplinary approach was indicated. 52 Using an interdisciplinary approach, combining theory and practice by incorporating the conducting of interviews with legal practitioners and a file search, provided a solid basis to draw up an in-depth overview of the legal practice with regard to victims' compensation. As mentioned, applying an interdisciplinary approach does not dispense with the need for a normative perspective as the latter is inextricably linked to doctrinal legal research, 53 since the object of the study is normative. Notwithstanding the overall evaluative nature of our research, leading towards evaluative answers, there was also a need to include a normative question. 54 Although we did expect that the solutions found would vary in their nature, e.g. institutional or policy based, it was clear that legal solutions would lie at the heart of the research. Our hypothesis, that was based upon previous evaluations, internal knowledge and practical experience, was that the research would indicate a need for (a further) adjustment of the law. Therefore a normative question was added to the central evaluative question, focussing on possible solutions that are present within the existing legal framework. Four legal arrangements were explored in order to assess their potential to solve the problems found in the evaluative part of the research.

Methods
The design of the evaluative study included three methods to answer the central question, which are: (1) a desk study, (2) file research and (3) semi-structured interviews. Using these methods, the research was split into five stages, stages 2-4 representing the focus of the study. A detailed description of both the design and the outcomes is provided in Sections 6, 7 and 8.
Overall the research design contains a combination of qualitative and quantitative methods, combined with a mono-disciplinary desk study. By combining different methods, data and expertise we expected to be able to provide an accurate picture of today's legal practice, confronting legal assumptions with practical findings. Pointing at the correlations within the findings would enable us to make a broad, yet in-depth analysis. Therefore we created a body of theoretical and organisational knowledge that was compared with contemporary legal opinions amongst legal practitioners (qualitative research), as well as with contemporary practice in terms of prevalence (quantitative research).
Each part of the research was analysed separately and if differences or inconsistencies emerged these were discussed amongst the researchers, as were the common features. Potential explanations for such findings, especially for the differences, were sought in comparing quantitative and qualitative data. Next, data relating to organisational aspects were taken into account. Subsequently, the findings (including the potential explanations) were discussed in various meetings with the supervisory committee.
Since the central question required the use of legal and social science methods, a team of researchers with different backgrounds had to be set up. Next to expertise in tort law and criminal law, there was a need for input from social sciences. In light of this, two social scientists (both criminologists) joined the research team. As mentioned above, this interdisciplinary co-operation turned out to be quite an endeavour in terms of clear communication. 52 One can also classify our research design as a 'mixed approach'. Leaving aside the issue of the preferable terminology, it is clear that the methods we used were chosen in light of the aim of our research: to analyse the practical application of the law on victims' compensation in criminal proceedings. 53 We use the term 'doctrinal' as applied by Bodig, supra note 6, pp. 47-48. Bodig uses the terms 'doctrinal knowledge' and 'doctrinal scholarship' to point to practice-oriented research born out of 'an awareness of interpretive variability, tensions and patterns of contestation', featuring the addressees and the officials as active agents (p. 48). The former implies that the normative perspective of the law is -to a certain extent -fluid and subject to fragmentation, since it is used as a vehicle to justify individual victims' claims for redress that in turn need to be accessed by officials. Using his legal experience, the accessor 'translates' the facts into a normative, legal assessment. 54 See Bodig, supra note 6, p. 48: 'Doctrinal research is normative and not simply because of its objects are norms and their justificatory implications but mainly because it is adjusted to a practical orientation: accepting normative guidance from a given practice and being committed to maintaining it in its integrity.'

Execution of research I: mapping the criminal proceedings (including the preparatory phase)
The first part of the study involved a theoretical analysis of criminal proceedings on the basis of legal and policy documents and academic literature. A description was drawn up, taken from a full body of literature and policy documents. Sources were found via regular search machines. 55 Next we studied internal policy documents provided by the respondents.
The mapping of criminal proceedings also related to the preparatory phase. Although the explanatory memorandum of the Research and Documentation Centre did not relate to the preparatory phase, it was clear that the activities offered in support of victims' compensation arranged within the Victims Support Desk needed to be taken into account. Notwithstanding the right of the victim to lodge a claim at the beginning of the trial, the legislator expressed a clear preference for lodging the claim during the pre-trial investigation. Today Victim Support the Netherlands offers assistance to the claimant at an early stage of the proceedings, whereas formerly a final, formal check was carried out shortly before the trial. The current policy starts from the assumption that a timely preparation furthers the awarding of compensation. Policy measures, which had been adopted to rearrange the support provided within the Victims Support Desks, were of relevance to this research.

Design
By analysing legal documents, policy papers, academic literature and, where necessary to be able to understand the policy documents correctly, conducting semi-structured interviews, the practices within today's criminal proceedings were mapped. Semi-structured interviews in the context of this study means interviewing on the basis of a questionnaire with the flexibility to ask follow-up questions or to deviate from the questionnaire depending on the circumstances. The method of semi-structured interviews was chosen due to the fact that it accommodates flexibility, which allows specific issues to be addressed in more detail. 56 Using semi-structured interviews provided a rich source of information since the respondents did not solely confine themselves to information with regard to the preparatory phase, but made spontaneous statements regarding their appreciation of victims' compensation and the problems they observed with regard to both the pre-trial stage and the trial stage. This (additional) information was referred to in the interviews with judges, prosecutors and lawyers.
Paying a visit to two Victims Support Desks provided insights into the practices and problems that needed to be explored. After drawing up a questionnaire, 57 semi-structured interviews were conducted with 17 respondents working at several Victims Support Desks or who were involved on a policy level. 58 Having an awareness of regional differences with regard to policy and practice, limitations concerning means and time and the subsidiary nature of this part of the research led us to decide that we had to settle for an initial exploration, instead of an in-depth analysis. Purposive sampling was used to select the informants because this is a non-probability type of sampling in which respondents or groups are selected based on characteristics instead of being representative of the population. 59 The respondents were collected indirectly via contacts of the project manager and a member of the supervisory committee. In other words, future respondents were recruited from among their acquaintances and through the first point of access. 60 The 55 Since the search related to documents which, in turn, were related to Dutch practice, we used search machines that primarily relate to the national setting like Google Scholar and Picarta. Moreover, due to the expertise of the project leader many articles, reports and policy documents were already known to us. In order to enable the respondents to prepare for the interviews a questionnaire was sent beforehand. Other than may be expected because of the semi-structured nature of the interviews, we did not use a topic list. We did not do so because we wanted to gather as much information as possible with limited means and time, since an exploration of the pre-trial stage did not form a part of the original research questions put forward by the Research and Documentation Centre. Moreover, we were aware of the regional differences within policy and practice and did expect to be confronted with distinct opinions and practical experiences. 58 We were able to arrange interviews with respondents working on behalf of 6 (of the 9) regional Victims Support Desks. majority of the interviews took place face to face, although some interviews were conducted by telephone due to time constraints. The interviews were recorded and later analysed by the project manager.

Results
The most significant issue mentioned with regard to the pre-trial stage was the risk of the non-registration of the victim in his or her quality as an injured party. This is said to be caused by the digitalisation of procedures and the short timeframe which is legally available for filing a claim. Respondents, however, mentioned 'work around' solutions, which enabled them to restore omissions, given that these were spotted in due time. Nevertheless, the ongoing digitalisation was also valued. Indeed, non-registration can also be caused by inactivity on the part of the injured party. Problems experienced by those working at the Victims Support Desks are dealt with by applying 'work around' solutions, emphasizing the importance of personal contact. Nevertheless, non-registration and other administrative omissions in the pre-trial stage regularly cause problems at the trial stage.
Other issues mentioned concern the time pressure and a lack of capacity at Victim Support the Netherlands, implying diminished assistance with regard to the claims lodged by injured parties. Moreover, respondents other than those from Victim Support the Netherlands suggested that Victim Support the Netherlands needs to lower victims' expectations; an argument that is subscribed to by the courts. Here lies an important difference of opinion between Victim Support the Netherlands and lawyers, on the one hand, and the Prosecution Service and the courts, on the other: whereas the former adopt a civil-law perspective and pursue tailor-made compensation, the latter tend to assess the claim from a criminal lawlike perspective, implying a more general approach using 'fixed rates'. 61 Similar arguments were heard in the evaluation in 2007. Overall, the conclusion is that although the present preparation of victims' claims is still not trouble-free, the current policy is preferable to the model used in 2007.

Execution of research II: the quantitative research
The quantitative part of the research involved two activities: analysing authorized figures for the period 2010-2014 and the execution of a file search during 2014.

Authorized figures
Since there was a relevant legislative amendment in 2010 (the introduction of a new admissibility criterion, i.e. the aforementioned 'undue burden' criterion), we opted for the period 2010-2014. 62 In general it is rather difficult to find authorized figures in the Dutch setting; only a few databases can provide detailed information as required for the kind of research carried out in this report. The RAC min database, a major database managed by the Research and Documentation Centre, section SIBA, is the best for these purposes. 63 However, since RAC min is a large database with underlying particular systematics, coming up with an instruction as to which data need to be selected is quite an endeavour. Moreover, we learned that the classification of RAC min does not always correctly correspond with the legal issues addressed. For example, the registration rulings of a compensation order (Article 36f DCCP) are registered on a case level, but the ruling does not have to apply to all the crimes assessed. In our report we give an account of the corrections applied.
Based on the authorized figures taken from RAC min we found that the total number of criminal cases in which a claim was put forward by an injured party remained relatively stable, with an average of 17,909 cases annually. The figures show a tendency towards a decline in the number of cases found to be wholly inadmissible. Indeed, the number of cases found to be wholly or partially admissible had doubled in five years from about 30 percent in 2010, to 60 percent in 2014.

File search
For the purpose of the file search, we selected four categories of crimes: property crime, simple assault, vandalism and public violence. We expected these categories to generate 'manageable' civil claims that fit the admissibility criterion. It would, in other words, in principle be reasonably possible to allow a civil claim in these types of cases. In these cases it is thus particularly interesting what factors preclude admissibility to subsequently see whether these obstacles can be eliminated. Next, we opted for files handled by a threejudge panel (meervoudige kamer) of the court of first instance since these panels provide for an extended ruling, whereas the single judge (politierechter) only provides an oral ruling. The year 2014 was chosen in order to ensure that the ruling had become final at the time of the file search.
To prevent an excessive burden for the organisations involved, the number of files was limited to a total of 180. Since the aim was to present a cross-sectional overview, 6 regional courts were selected and this choice was made because they, based upon RAC min, had dealt with a substantial amount of cases containing civil claims in 2014. The courts selected were regionally spread out representing both rural and non-rural areas. For each location 30 files were selected, plus an additional 10 to cover fall outs. 64 The cases were randomly selected via RAC min. Random sampling is a probability sampling method that relies on the law of probability to select a subset of a statistical population in which each member of the subset has an equal probability of being chosen. 65 Similar to the authorized figures, problems arose because of inconsistent registration. This caused, amongst other things, a fall out of files because not all the files selected contained useable information. For example, cases that were registered as containing a ruling on compensation were actually related to other categories of crime than those selected for our research or where an acquittal was ordered. In the end, out of the 240 files that were studied (all additional files were studied), 217 useable files emerged. This had repercussions for the authority of the findings. For the purpose of the analysis of the files a list was designed consisting of 131 variables 66 that could affect the admissibility decision, and they were placed in the chronological order of the judicial decision-making. The list included, for example, the nature of the damages and the extent of the claim. The list was taken from the evaluation of 2007 and modified. Some items were left open in order to provide room for scoring variables that were not mentioned in the topic list.
In order to guarantee the consistency of the file search conducted by the distinct raters, the first ten files were analysed in couples. The files were rated by the legal experts, except for the first two sessions at which a criminologist was present, in order to provide consultation with regard to the use of the database. This inter-rater reliability was performed to test consistency among observational ratings provided by multiple coders. 67 All raters used a journal to note apparent inconsistencies and peculiarities in the files, and these findings were discussed during the rating sessions and the outcomes were also noted in the rater's journal. Moreover, the project manager was present at all sessions in order to further a consistent execution of the sample study. Where inconsistencies were found in the initial statistical analysis, the (journal of the) rater concerned was addressed and corrections were made. In the end a final plenary control of the database took place so as to detect inconsistencies and uncertainties. 68 The data were recorded in Excel and were later imported in SPSS 21.0 in order to be subject to statistical analysis. Besides descriptive statistical analyses, a Chi-squared test (X2 test) was used to determine whether a significant difference occurred between expected frequencies and observed frequencies; 69 for example, to establish whether there were distinct outcomes between different groups (e.g. whether an injured party made (no) use of legal assistance). The statistics from a chi-squared test arise from the assumption  70 An ANOVA test (F statistic) was used to compare the ratio of variance between groups and within groups (e.g. the amount of loss claimed). 71 With regard to variables representing a substantive variable (e.g. the amount of the pecuniary damages claimed) both bivariate regression analysis and a coupled T-test were applied. 72 A T-test was used to test the differences between means. 73 For instance, to measure the mean effect of the prosecutor's opinion with regard to the amount of compensation to be awarded with the compensation awarded by the court. Finally, the significant variables were grouped to perform multivariate analyses. By making use of a multivariate analysis, the significant effects of the independent variables can be tested simultaneously when these are placed in a single model. 74

Results
The statistical analysis showed multiple results, which cannot be mentioned in the context of this article. Therefore, we will single out the most significant ones. Beforehand, however, we must give an account of the list, which could be considered to be too detailed, and its rather rigid structure. Basing the design on the chronology of the decision-making within the criminal trial turned out to be somewhat inadequate since it gave way to repetitious variables. With the benefit of hindsight it is clear that the list of variables should have been tested more extensively beforehand.
The file search, an inquiry into four categories of crime, provides a similar picture as mentioned for the authorized figures. The indication that points to a substantial decrease in the amount of cases that were found to be completely inadmissible was established. Only less than 1 out of 10 claims was found to be completely inadmissible. The vast majority of the claims analysed were found to be admissible (entirely or partially), so that compensation was awarded. Most of these claims related to pecuniary damages (7 out of 10 cases), whereas non-pecuniary damages related to 1 out of 4 cases. With regard to their number and nature (pecuniary or non-pecuniary damages), the outcomes were not discriminative in nature, albeit that the multivariate analysis shows that the success rate increases if the claims relate to non-pecuniary damages exclusively.  What attracts attention is the adjustment of the extent of the damages. Both the prosecutor and the court show a tendency to lower the amount of compensation. Based on these files, the average amount of pecuniary damages awarded was 2,619 euros; with regard to non-pecuniary damages the average was 905 euros. There is a significant relationship between the quality of the claimant's arguments and the awarding of compensation for damage suffered by the injured party. Moreover, claiming pecuniary damages for multiple items of loss has a positive effect on the awarding of compensation. It needs to be mentioned, however, that the majority of the claims related to property crime. The items of material damage posted were mostly the loss of material objects or the loss of income. Using the standard-form spread by the Victim Support Desks also shows negative effects on admissibility. With regard to non-pecuniary damages, legal assistance shows a significant positive effect on admissibility. In about half of the files, objections raised by the defendant/tortfeasor resulted in a ruling of partial inadmissibility. Finally, if the claim for damages is allowed, the compensation order is applied.

Execution of research III: the semi-structured interviews relating to the trial phase
Since the legislator tends to refrain from providing clarifications with regard to civil compensation in the context of criminal proceedings, in particular with regard to the interpretation of the recent admissibility criterion (i.e. undue burden; onevenredige belasting), it is up to the judiciary to interpret these provisions. This means that the judge's opinion on the admissibility criterion is of major importance. Notwithstanding the instruction by the Council for the Judiciary, the professional, albeit personal opinion of the judiciary has a substantial effect on the outcome of the claim's assessment. This was also one of the findings in the 2007 evaluation. However, developments have rapidly occurred since then: next to the extension of the admissibility criterion, the judiciary nowadays has to deal with extended victim participation and victim advocacy has been on the increase. This raises the question of whether these (political) legal changes influence the decision-making by the judge, especially with regard to the admissibility ruling. Notwithstanding the existence of an extensive body of jurisprudence on this issue, there is still room for interpretation on a case by case basis, implying a need to conduct semi-structured interviews to discern the law in action.

Design
To gain an insight into the judiciary's considerations, semi-structured interviews were arranged. Beforehand, an in-depth desk study on the situation with regard to the legislation and jurisprudence was conducted. Literature was taken from regular databases, such as Picarta, World-cat and Web of Science; the case law was extracted from the national database (Rechtspraak.nl) and the Dutch Journal on Jurisprudence (Nederlandse Jurisprudentie). Several search terms were applied relating to victims' compensation, e.g. undue burden (onevenredige belasting), compensation (schadevergoeding) or compensation order (schadevergoedingsmaatregel). Extensive use was made of the official advice on how to deal with victims' issues, published by the Council for the Judiciary in 2013. 75 The respondents were selected by the Council for the Judiciary, the Public Prosecution Service 76 and, with regard to the lawyers, they were selected from the network of one of the researchers. The courts that were approached to invite judges to participate in the interviews were regionally spread out, like the courts that participated in the file search, from the same rural and non-rural areas. The courts carried out the ultimate selection of the respondents. This means that it could well be that the respondents had strong views on the topic, either positive or negative. However, judges are highly intelligent professionals and are well aware of their personal attitude to a topic. We assumed that these respondents could effectively distinguish the general picture on the interpretation of the admissibility criterion in court from their personal view based on their own experience. Moreover, we included a particular question on the respondent's perception of 75 J. Candido et al., Slachtoffer en de rechtspraak (2017). 76 The respondents from the Public Prosecution Service were invited by one of the members of the supervisory committee. Two of these respondents were from a second instance court and were Advocates General (advocaat-generaal); the others fulfilled the function of the public prosecutor at the first instance (officier van justitie).
Utrecht Law Review | Volume 13 | Issue 3, 2017 | Special Issue: Methodology of Legal Research his colleagues' attitude towards civil damages compared with his own. The same approach was applied with regard to respondents from the Public Prosecution Service. Since the Council for the Judiciary was only willing to select 8 respondents, the same number of respondents was also chosen for the other organisations in order to balance the outcomes. In the end 24 interviews were conducted: with 9 judges, 7 public prosecutors and 8 lawyers. A questionnaire containing 33 items was sent beforehand to be able to get the most out of the limited time that was available for the interviews. Different questions were formulated for the three kinds of respondents. However, the distinct questionnaires were almost identical. In order to further a comparison of opinions, the three different questionnaires contained the same case study, which all respondents were invited to assess. The interviews were recorded with permission 77 and typed out.
The scripts were analysed in NVivo 78 by two researchers. In order to establish inter-evaluators' trustworthiness, several reports were first analysed in pairs. After having decided upon the relevant textual elements ('nodes'), both researchers individually made an analysis of the scripts. 79 These analyses were subsequently discussed and any question marks that arose were answered by taking a second look at the scripts. Subsequently, the concept chapter that was drawn from this analysis was presented to the other researchers who had participated in the interviews. 80

Results
There are clear similarities between the findings in this study and the ones assessed in the 2007 evaluation, albeit that no lawyers had been interviewed back then. Problems with regard to the inadmissibility of insufficiently founded claims, or the need to lower the amount of damages awarded, were still mentioned. In both studies the 'need' for victims' compensation is commonly accepted amongst the judiciary. To date, victims' compensation has become a regular element within the criminal procedure. This apparent willingness to provide victims with compensation does not imply that the criminal justice authorities are without criticism. Judges and public prosecutors relate the issue of victims' compensation to the overall challenge -if not the struggle -of upholding the rule of law. Victim participation has become increasingly intrusive, thereby posing a challenge for judges and prosecutors when holding a proper court session. In light of this, some judges, albeit there are only a few of them amongst the respondents, are of the opinion that victim compensation needs to be dealt with in a private law context.
The objections mentioned appear to resonate in the awarding of claims. The majority of both the judges and the public prosecutors stated that claims tend to be overrated and are in need of adjustment. Being confronted with the outcome of the sample study that the compensation awarded is often less than requested, one is not surprised. With regard to the nature of the damages and their amount, both are not considered to be problematic in light of the undue burden admissibility criterion. The turning point lies in the quality of the claim; if the claim is properly substantiated, compensation will be awarded. A substantive cohort of claims, however, is found to be inadequately substantiated. Whether or not the claim will be awarded depends on the overall profile of the case, the level of substantiation and the profile of the judges, especially the amount of civil-law expertise available at the court in question. Remarkably, lawyers prove to be the most critical with regard to the need to substantiate the claims, but they tend to start from a civil law perspective. So does Victim Support the Netherlands. However, disappointment lies ready and waiting since verdicts finding the claim to be partially or wholly inadmissible, or ordering an adjustment of the claim, are presented without a clear explanation. According to Victim Support the Netherlands and the lawyers this evokes secondary victimization, especially when the public prosecutor requests the adjustment. Both the judges and the public prosecutors are aware of the risk that the victim will be disappointed, but reject responsibility for this. Given the inherent limitations that go hand in hand with the public nature of criminal

Execution of the research IV: a SWOT analysis of the legal figures
Obviously, the empirical parts of the research show that progress has been made. The increase in the amount of claims (partially) awarded shows that victims' compensation in criminal proceedings is not about getting all excited over nothing. On the other hand, however, the findings do show that practical problems remain, and that principal objections are still present. Indeed, the judiciary struggles with the challenge of how to uphold the rule of law in light of extended victim participation. Since the Dutch criminal justice system features a non-bifurcated procedure it is difficult, if not impossible, to reconcile the defendant's claim to have the crime assessed in light of the presumption of innocence and the victim's simultaneous claim to be compensated for the crime (seemingly) committed. Moreover, today's criminal policy is based upon an efficient handling of criminal cases, including potential claims for compensation, leaving no room for postponement, e.g. to provide the victim with an opportunity to substantiate his/her claim.
The aim of our research, however, exceeds such a description of the state of the art: the aim was (also) to contribute to the search for (legal and practical) opportunities to further civil compensation on behalf of crime victims The empirical part of the research served to draw up an inventory of contemporary practice, the problems present and the solutions found. In order to deliver a significant contribution to the legal debate on how to further crime victims' compensation, these empirical findings need to be related to the law's procedural design in order to analyse whether the present provisions contain opportunities that could potentially further civil compensation, if needed by means of minor legal amendments. For this reason we conducted an additional theoretical research.
Taking the present system represented in the DCCP as the point of departure, the potential of four legal arrangements that could satisfy the demands of the different actors involved was explored. Each arrangement was made subject to a SWOT analysis (Strengths, Weaknesses, Opportunities and Threats analysis), starting from four different scenarios. The scenarios refer to the procedural position taken by the defendant towards both the crime and the claim. Next, we developed items that could serve to assess the strengths and weaknesses of each legal arrangement in the setting of the distinct scenario, e.g. the desire to optimize the victim's compensation and the need to uphold the presumption of innocence. Drawing up this rather detailed and complicated design was done by four researchers, all legal academics. The framework of analysis was discussed extensively in several plenary sessions. After having found a consensus on the design and the interpretations to be used, concept analyses regarding the different legal arrangements were individually made. These concept analyses were later discussed in plenary session, in order to guarantee the objectivity and consistency of the SWOT analysis. Thus a careful analysis of the potential of the selected legal arrangements was made possible and that can serve to deal with the theoretical and practical problems that were assessed in the empirical part of the study, providing food for thought as to whether there is a need for legal change.
This part of the research was extensively discussed with the supervisory committee. However, there was no intensive discussion with the social scientists from the research team. They were informed about the outcome of the legal analysis, but did not participate in the process due to the exclusively legal nature of this part of the research. This part of the study was therefore not subject to triangulation.

Evaluation: opportunities and difficulties in conducting empirical-based legal research -the case of triangulation
Since this study relates to an evaluation of a predominantly legal practice, i.e., victims' compensation through criminal proceedings, it stands in 'the shadow of the law'. In line with this, and in our opinion, a legal evaluation concerning the 'law in action' is inextricably related to 'the law in the books'. The aim of our research was to map the opinions of legal professionals and the problems they encounter with regard to victims' compensation. Obviously, this entails implications for the nature of the research and its methodology.
As is the case for all legal research, hermeneutics are implied to further the understanding of the law on both a theoretical and practical level. 81 One has to map the arguments that underlie the law's application by legal professionals. 82 This does not suffice, however, because in an era of pursuing socially effective judicial adjudication legal practice also needs to be evaluated in terms of its legitimacy. This calls for interdisciplinary research which aims to answer evaluative questions that imply the use of methods that are currently not yet used by legal researchers, but are indeed used in social sciences in particular. Obviously this enriches but also complicates legal research. When faced with the challenges of conducting interdisciplinary research (with multidimensional aspects being added) within a limited timeframe, we opted for triangulation. 83 This proved to be a good decision as we had to face multiple difficulties along the way. With regard to external factors, these were of a varied nature. Some hurdles which we had to face could be just a matter of 'bad luck', such as delays as a result of miscommunication with the courts' administration and difficulties in contacting the persons involved. The subsequent execution of parts of the research were hindered by bureaucracy, the limited timeframe of the research implying that one has to be flexible when conducting several parts of the research. Nevertheless, once permission to study the files and to conduct the interviews had been given, a warm welcome and support awaited. This was the same for communication with the SIBA section, the administrator of RAC min. 84 There were, however, other potentially threatening external (f)actors that were related to the availability of the quantitative data. Fundamental issues needed to be solved with regard to RAC min. Since RAC min is a huge database with a distinct system, requests for data needed to be tailored towards the classifications underlying RAC min. The latter, however, did contain (legal) inconsistencies. Moreover, since our contact came from a non-legal background this added to the difficulties in selecting the data and the files required. As a result of the aforementioned legal inconsistencies, some of the files selected proved to be unusable. Moreover, the sample study was limited to four categories of crimes. Here, triangulation proved its value. By linking the quantitative outcomes to the findings assessed in the qualitative part of the research we were able to detect the common denominators that served to preserve the consistency of the outcomes. 85 Other parts of our research did not prove to be a problem. A desk study being the core business of legal academia, this did not result in any major issues. Still, obtaining policy documents that provide a picture of the practices which are present at the pre-trial stage sometimes proved to be difficult since these documents are not always open to public scrutiny or are difficult to find. In our case, the respondents interviewed with regard to the pre-trial stage were very supportive; they provided us with policy documents and relevant information. But exploring the pre-trial stage did add an extra layer to the research since it was absent in the preliminary memorandum. Here one may also question whether we exceeded the legal domain and inadvertently entered the domain of governance. Contemporary legal researchers are used to conducting qualitative research, but this is subject to a caveat as they might lack the methodological capacities to do so properly.
Moreover, we had to face another potential 'weakness' since the number of interviews regarding the trial phase was limited. Conducting the interviews did not result in any obstacles as these were carried out by the legal researchers, under the supervision of the project manager who has ample experience in (the methodology of) conducting and assessing interviews. Conducting interviews with different types of respondents, using similar questionnaires, including a case study, provided us with the tools to discern a consistent pattern of decision-making and underlying (differences in) opinions amongst legal practitioners. The finding that the outcomes of the interviews matched those of the sample study provided confirmation of the adequate nature of the image of legal practice as reported by us. Again, the decision to opt for a combination of research methods (triangulation) proved to be a good choice.
As regards the approach of triangulation more generally, it can be concluded from our research that the starting point of answering one and the same question while using different methods in practice works out somewhat differently. It is clear from our study that the empirical data that were gathered by both quantitative research (a file search) and qualitative research (interviews) are in need of (legal) interpretation. As argued by Taekema & Van der Burg, a particular legal discipline entails distinct methods and objects of research, as well as a distinct conceptual framework. 86 Thus, when conducting legal research one cannot -and more importantly, should not -depart too much from the law's hermeneutics. Especially not if one uses, as is the case for our research, a multidimensional approach next to an interdisciplinary one. The former, however, challenges the law's hermeneutics as an encounter between the law and, for instance, the social sciences evokes a need to find 'common ground' in terms of the object of the research, the terminology used and the methodological conditions to be met.
Moreover, empirical data gathered on behalf of the judicial system have inherent limitations. On the one hand, setting up a registration system such as RAC min implies that there are institutional interests involved, indicating that some data are registered, whereas others are not. Moreover, as we experienced, such registration may contain categories that do not comply with the legal qualifications and/or may not be related to the individual offences registered. The qualitative data collected from interviews, on the other hand, tend to reflect the professional, yet personal opinions of the officials involved. The institutional background, the position and the personal experiences of the distinct respondents determine the outcomes of the interviews. Since the institutional background and the activities of the distinct respondents differ, the answers will relate to different perspectives and various situations. Such a variety of information inevitably evokes a need for interpretation by the researcher. As argued, the execution of interdisciplinary research implies that one is aware of the need to develop common terminology, or at least to avoid potential errors in communication. Indeed, triangulation is not the solution to the problem, but its potential cause. Opting for interdisciplinary research therefore also requires 'researchers' triangulation', implying the need to create a mutual understanding or at least to create awareness with regard to the potential for unconscious miscommunication.
The former relates to the challenge in facing internal differences that come along with both an interdisciplinary and a multidimensional approach. Both call upon the research team to overcome the challenge of co-operation and the related need to create a mutual understanding. First, both criminal and civil lawyers needed to co-operate when facing the endeavour to study tort and crime in an era that is full of victims' participation and claims for compensation. We were challenged to explore the traditional concepts of tort and crime, paying tribute to the legal convergence that is present and the changing image of liability law that flows from these developments. We provided the body of knowledge that was needed to assess current legal practice. In fact, we had to adapt to the diffuse (legal) reality of (preparing and awarding) victims' compensation and to become familiar with the problems faced by respondents on a daily basis.
An even bigger challenge was the triangulation of expertise, i.e. developing a mutual understanding between the lawyers and the social scientists in the research team. This holds true for both methodology and terminology. In general lawyers are not known for their methodological or their statistical knowledge. The object of legal research being the consistent design and application of the law, legal research traditionally tends to be of a one-dimensional nature. 87 Moreover, the methodological standard applied by lawyers tends to be of a rather intuitive nature, using different methods and sources than those applied by social scientists. 88 Conversely, social scientists are not used to operating in a deeply normative context such as the law. They do however apply strict rules with regard to methodological accountability, especially with regard . 87 This is not to say that legal research is of an applied nature since its aim is to further the relevance of the use of the law in legal doctrine and serves as an instrument of social engineering; see Herweijer, supra note 30. 88 See Herweijer, supra note 30, para. 5. to quantitative research. Getting a grip on both worlds in order to further a mutual understanding of both the meaning of the judicial items tested and the results of the quantitative study proved to be a challenge. This led to repetitive consultations resulting in the secondary processing of data and an internal review, especially with regard to the quantitative data. 89 Here we end our report on a complex, yet enjoyable methodological journey. The interdisciplinary design requested in the memorandum by the Research and Documentation Centre proved to be a methodological endeavour, but one which we successfully faced and -more importantly -one that we enjoyed and are determined to pursue. Crossing the internal and external boundaries of the legal domain in order to conduct empirical legal studies provides for mutual learning and understanding. Moreover, among legal researchers and practitioners it raises awareness of the need to account for the somewhat implicit assumptions with regard to the law's legitimacy that are present within the current legal discourse. Indeed, legal theoretical research is of great importance, but since the law is increasingly called upon to serve as an instrument to steer social processes, legal research should also open up to study 'the real world'. 90 Applying triangulation provides opportunities to reflect on the legal position that accompany interdisciplinary and -to a lesser extent -multidimensional research. It provides a vehicle to preserve (at least to a certain extent) the overview that is traditionally longed for by lawyers and provides them with guidance in finding their way through the 'jungle of interdisciplinary research'.
The former, however, does not provide a solid answer with regard to the question of how to genuinely overcome the difficulties that lawyers (or for that matter other academic specialists) face when they opt for an interdisciplinary approach. It feels like our journey has just begun, but by using the triangulation approach we did find a way to navigate the rough waters of interdisciplinary research. What is more, some may argue that we did not even reach these waters and that our research should be qualified as monodisciplinary, with 'an interdisciplinary touch'. Given our serious efforts to confront the challenge of both cross-border legal research (multidimensional) and cross-border discipline research, by finding relief in triangulation as an approach to find common ground, we respectfully reject this (dis)qualification. 91 Triangulation is certainly not a wonder drug with which we can solve fundamental methodological issues, but it is a useful vehicle to get a grip on the methodological difficulties that accompany legal research with regard to the 'law in action'. Moreover, we cannot disengage from the normative nature of the law, since this is the reason for the law's very existence. A similar argument can be made concerning any other academic discipline, since the reason why one can claim that it is a 'discipline' is its internal consistency and methodology. Finding an answer to the question of whether and how we can develop normative legal conclusions on the basis of multidisciplinary or interdisciplinary research implies that one pays tribute to the law's uniqueness. 92 Therefore, we need to progress with cat-like steps into the domain of the challenging discourse of legal interdisciplinary methodology in social reality. 89