IASB ED: MANAGEMENT COMMENTARY: An analysis of experts´ opinion through a delphi method Igor Álvarez (Basque Country University) José Antonio Calvo (Basque Country University) Araceli Mora (University of Valencia) 1 1. MOTIVATION, OBJECTIVE AND METHODOLGY In June 2009 the IASB Exposure Draft “MANAGEMENT COMMENTARY” has been issued with a deadline for comments by 1 March 2010. In order to give a relevant input on the issue to the IASB and to the EFRAG (TEG) we have addressed a research based on a “Delphi model” which have been increasingly used in Social Science to deal with problems (including normative matters) for which consensus among experts can help to policy makers. We briefly explain below how the method works. Although we give a complete explanation and literature review in the annex. Delphi method is a scientific and rigorous method of analysis widely used in social sciences in the last decades. According to the classical definition, the Delphi method is a general way of structuring the group communication process and making it effective enough to allow a group of individuals, functioning as a whole, to deal with complex problems. It is a systematic process, which attempts to obtain group consensus resulting in much more open and in depth research than a traditional survey. The individuals involved are a small group of experts in the topic (known as the Delphi panel). The structure of the group is designed by a monitor/s (the researcher/s) that formulates a reiterative survey to address the research topic. The survey is sent to the designated group of experts who subsequently return their responses to the Delphi monitor. Each of the iterative mail-outs of the survey is called a round, and rounds continue until stable responses between rounds are achieved (usually two or three rounds are enough). The Delphi does not depend on statistical power, but rather on group dynamics. A key advantage of the approach is that it avoids direct confrontation of the experts (as they do not know who are the others during the process) while they give their views and arguments. This allows richer argumentation and better validation of arguments. It has the advantages of research techniques based on group interaction, but ensuring that individual answers are anonymous for the panellists and, therefore feedback is unbiased as to the source of the original opinion. So, it documents the opinion of the panellists, while avoiding the pitfalls of face-to-face interaction such as group conflict and individual dominance. At the same time controlled feedback and anonymity can help panellist to review their views without publicly admitting that they have done so, encouraging them to take up a more personal viewpoint rather than an institutional position. 1 Address for correspondence: Araceli.Mora@uv.es The panel of experts is divided into groups with different backgrounds, which permits comparison of the perspectives of the different groups. The results of the Delphi method do not try to substitute the traditional comment letters but might complement them. Comment letters are generally sent by organizations representing the “institutional opinion”, as well as regulatory bodies. Due to different reasons, individuals and non- institutional opinions (as for example from the academics) are less probable to participate in the due process, as they do not normally send comments letters, so interesting views are lost in the process. The Delphi method gives another approach as it results from the exchange of arguments among the commentators, and it allows an analysis of this anonymous exchange of arguments with a scientific and rigorous methodology which permits avoiding a great part of the subjectivity inherent to any analysis of opinions. This research is focused on ED Managemnt Commentary with the intention to help policy makers in their decisions. In this study we consider, among others, the main questions related with the ED, including the specific questions to constituents, and show the opinions and arguments of the selected experts and their ex-change (and change) of views through the process. Next we indicate the experts panel (names and affiliations), them we show the main conclusions of the study in relation with the different topics in the ED, and finally we explain in more detailed the research process and the statistical results. 2. THE EXPERTS PANEL The panel consists of 22 experts carefully selected from different European countries (UK, Germany, Austria, Sweden, Spain, Switzerland, Denmark, Italy). Among them, there are two main groups: a) Non-Affected: - Prestigious European academics involved in research related with financial reporting and disclosure and, in most cases, having been involved in some way in the national or international accounting regulations processes. b) Affected: - Auditors (partners in the big firms) in charge of the IFRS matters - Managers from big listed European companies - Managers from big European investing societies. The name of the experts and their affiliations are shown in table 1: Panel A: Academics MANFRED BOLIN LEANDRO CAÑIBANO ROBERTO DI PIETRA GÜNTHER GEBHARDT BEGOÑA GINER JOSÉ ANTONIO GONZALO JAN MARTON FLORA MUIÑO FRANK THINGGAARD ALFRED WAGENHOFER GEOFFREY WHITTINGTON Pane B: Affected (preparers/auditors/users) CLEBER CUSTODIO ALAN DANGERFIELD ARANTZA GINER JESUS HERRANZ/ROSALIA MARTINEZ JORGE HERREROS SIMON INGALL PAUL LEE ROBERTO MONACHINO MARTA SOTO MIKE STARKIE CARSTEN ZIELKE Table 1: Panels of experts Institution International School of Management, Dortmund Autonomous University of Madrid University of Siena Goethe University of Frankfurt University of Valencia Alcalá University Gothenburg University Carlos III University Aarhus University University of Graz University of Cambridge Organization DELOITTE Roche BDO Ferrovial KPMG Shell International Hermes Equity Ownership Services Unicredit Telefónica British Petrolium Société Générale 3. SUMMARY OF THE FINDINGS AND CONCLUSSIONS Two rounds were enough to get stability in the answers. That means that the experts did not change their responses significantly from the statistical point of view, with a couple of exceptions. There are interesting qualitative changes in some of them, in the additional arguments. The most interesting part of the second round was the increase in the number of arguments in favour or against the positions taken by some respondents in the first round. The higher changes from the statistical point of view are related with qualitative characteristics and with the list of contents. The questions addressed in the questionnaires are shown in table 2, and the statistical results in both rounds are showed in table 3. Before showing the research process and the results we can summarize the main conclusions of the study in the following paragraphs: a) b) Usefulness and users of MC Everybody agree the MC is useful to make decisions as it provides an input to put the numbers of the financial statements into the proper context, although many respondents do not give the highest mark (5. Strongly agree) as they point out that usefulness can be lost in practice by the (low) quality of the MC presented by the companies, and some others agree about the “potential usefulness” subject, in some way, to its final structure. Everybody agree that providers of capital (users and creditors) are the primary users of MC, althoughany experts stress the importance of other users, mainly employees, and at a lower level competitors and institutions. c) Mandatory versus non-mandatory status The great majority of the experts, independently of their background, agree with having a non-mandatory guide “at this stage”. However, it is important to remark that this is one of the responses that have slightly changed in the second round due to the feedback with the arguments in favour and against the different status and considering the possibility of deferring the decision about the status in the next future. While most of the auditors and preparers insist on having a long term non-mandatory guide, some other experts (4 academics, 3 preparers and 1 user) consider the convenience of having a non-mandatory guide now, while experimenting the usage of the MC, and in a short/ medium term period it would be possible to evaluate the decision to make mandatory its preparation. Just 1 expert (academic) consider the MC must be mandatory at this stage. It is important to point out that there has been a change in views in this second round as a consequence of the feedback and now more experts show doubts and indecision in this matter giving a “3” (neither agree nor disagree) or a”4” (agree with IASB decision) but introducing some comments about that potential change of the status in the future The arguments of most of the auditors and preparers for having a long-term nonmandatory guidance are that a) will more likely avoid unnecessary contradictions in some jurisdictions while helping enhance reporting practices and provide a useful reference framework in jurisdictions that lack established reporting requirements of a MC-type and b) pointing out in some other cases that the management commentary are beyond the financial reporting, and that is not a matter to be considered by IASB (it is a matter of other institutions) or at least it is not a priority for the IASB. However, on the other side, half of the experts (in the second round) consider the IASB must (or should consider the possibility) issue a mandatory standard at some point in time to get a useful document .Their argument is that a non-mandatory guide does not get comparability and usefulness for the investors. They consider that a mandatory principles-based standard gives entities enough leeway to further improve it through voluntary disclosures and it does not contradict national rules, and the more principlesbased it is the less it generates a checklist to tick. It can be said that the experts, after reading the other people arguments, are clearly divided (around 50%) about the convenience of making the MC compulsory in the future. Although there is not a clear distinction among the backgrounds it seems that all the auditors get a high consensus being strongly reluctant to have a mandatory standard, while just 2 of the 11 academics consider the guide must be never mandatory. The overall conclusion can be that there must be a non-mandatory guide at this stage, although it seems to be a strong conflict of opinions about next steps d) Qualitative characteristics There is a great consensus about the importance of relevance, faithful and timeliness. Also there is quite consensus about the importance of comparability over time and understandability (although this last characteristic is related with the plain language in question 5.c). Although it seems to be low consensus in the question related with the use of plain language, after reading the arguments it can be concluded that most of the experts consider that the MC must be understandable, but always considering that it is addressed to “sophisticated users”. The main conflicts with the qualitative characteristics in the conceptual framework seems to be related with neutrality, comparability between companies and verifiability. We find the highest deviation in the opinions in the case of neutrality and comparability However it is important to remark that the mode is “4” meaning “very important” in both cases in the first round but “3” meaning “average” in the second round Neutrality: Although many experts consider neutrality has not sense in a document based on the management’s view on not only what has happened, but also why management believes it has happened and what management believes the implications are for the entity’s future. Some others experts consider that it depends of the real meaning of “neutrality”. Management could be expected to be transparent and provide its views in a balanced manner by including any relevant information in good faith, not having the intent of deceiving users, and presenting both positive and negative considerations and events. If neutrality definition is refocused in that sense then MC must have it. Comparability between companies: There is no consensus in this matter either. However, it is important to remark that only 6 of the 22 experts (with different backgrounds) gives low or none importance to comparability between entities arguing that it is very difficult to get it when giving the information through the eyes of the management. Other experts (7) consider that the same argument about difficulties in getting comparability between entities from different industries can be applied to any financial statement. Comparability between entities in the same industry must be a desirable characteristic of the MC and must be maximized (always considering the inherent flexibility of this kind of document). The rest has not a clear position, or think that it should debate with forthcoming Conceptual Framework. The case of verifiability is different as there are very few considering that it is not important, nobody said that it is very important. In fact the mode is 3 (average score). Many experts with different backgrounds point out that the characteristic of verifiability could be confusing. MC in practice, particularly if prospective information is included whereas users may consider important some sort of checking provided by an auditor, as for example consistency. Several responders show in the second round comments in relation with the importance of having more debate about the conceptual framework in general, including the qualitative characteristics, and pointing out this is not the proper context (the MC) to have such a debate. As an overall conclusion it can be said that there is a majority in favour of considering that “the desirable characteristics for MC must be the same that those definitively consider in the framework” (considering the aspects related with the refocusing meaning of some of them applied to MC, as neutrality or verifiability) and considering that more debate about the Conceptual Framework should be necessary e) Principled-based and list of contents There is a high consensus of agreement with the principled-based approach of the documents and at the same time with the list of contents showed in the document. However there are many comments in relation to the extent of some topics in the list, mainly in the second round, as for example: a) Results and prospects must be separated, and an analysis of variations over goals or prospective data provided earlier must be required b) More concrete information about risk and its management must be detailed in this list Also the extent of “looking-forward information” is a concern for some experts. The reaction of some experts to the feedback from the first round in order to consider other information as social, environmental or corporate governance makes some of them slightly changed their opinions by adding comments in relation of taking into account that information (social, environmental and corporate governance) to the extent this information is relevant to understand the strategy and overall performance of the entity. However, most of the experts tend to think that detailed information focused on the specifics of the environmental or social performance of the entity can be very useful and necessary, but a more appropriate content for specific separate reports. All the experts agree in having a principled-based document. However they also agree about having a list of contents being all the elements considered in the ED very important or essential (although a little bit more of detail about the content in some cases would help) f) Application guides and examples There is not consensus about this topic. While all the preparers and most of the auditors prefer not to have an illustrative guide or examples, the users are divided, and it seems that also the academics are divided in this topic. However, although 4 academics are in favour of not having a guide, the reason in two of these cases is that it is premature, but when getting more experience guides and illustrative examples would be convenient, and in one case the reason is that having a non-mandatory document with a detailed guide seems to be a curiosity. So, it could be said that the majority of the academics are in favour of having guides and examples now or in the next future. The reasons to be in favour are similar in all cases: a) The proposals in the exposure draft related to forward-looking information are difficult to put into operation without application guidance or/and examples. Examples would be helpful to get an idea how comprehensive disclosures are expected to be; and b) If the development of such application guidance and illustrative examples were left to other organizations (national or stock market regulators), there would be inconsistency and confusion – diminishing the comparability and usefulness. The reasons against having illustrative guides and examples given by the reluctant (mainly auditors) in the documents are that they might be interpreted as a floor or a ceiling. It is interesting that the only auditor in favor of having guides and examples argues that a very clear explanation that an example should not be considered as floor or ceiling can avoid that, and he/she believes that such risk is low compare to the potential benefits that an example can provide. As a conclusion it can be said that there is not a consensus about the convenience or not of having guides and examples, and this is one of the few topics in which we can find different opinions depending on the background of the experts g) The placement criteria Although with some concerns in relation with the importance of having a placement criteria, most of the experts agree with the decision to defer development of placement principles (where to disclose MC vs. Notes) until phase E of Conceptual Framework is completed. Although it is recognized in general the importance of the topic, considering that the intention of the IASB is to issue a non-mandatory guide, then the placement criteria could be quite confusing in those circumstances (although some others point out that it would be also confusing not having those criteria until the phase E is finished, maybe in some years) In conclusion it can be said that the majority agree with the decision to defer the development of placement principles, but pointing out that this situation can create difficulties in practice Other aspects: Finally some experts make comments related with auditing. Although this is not a matter the IASB has any competence, and so there is not any question related with that in the questionnaire. However this is an important topic that can affect the compliance and the usefulness of MC. Some experts consider the MC must be out of the scope of auditing as much of the information is prospective and subjective and cannot be verified (which affect the qualitative characteristic of verifiability). Others think that a requirement for an audit may encourage a rigid and excessively-cautious approach to management commentary. However, other experts point out that MC must be subject to audit of some sort, but taking into account the specific challenges of an audit of prospective financial information, stemming from its different nature. The auditor can merely assess whether the MC is” consistent” with the annual financial statements. 4. THE RESEARCH PROCESS The phases have been: 1. Selection of the experts 2. Sending the questionnaire in the first round 3. Qualitative and quantitative analysis of the responses 4. Preparation of the feedback report for the panellists and the questionnaire of the second round 5. Analysis of the final responses and elaboration of the report 4.1. The questionnaires The first questionnaire round has closed and opened questions, and it allows the respondents freedom to define the topics that should be included. We included all the questions addresses to the constituents in the ED. The close addressed questions and the scales are shown in table 1. Table 1: QUESTIONS ADDRESSED IN THE QUESTIONNAIRES Linkert scale from 1 to 5 meaning: 1. Strongly disagree, 2. Disagree, 3. Neither agree nor disagree, 4. Agree, 5. Strongly agree(except in Q.1.: 1. None knowledge, 2. Little knowledge, 3. Average knowledge, 4. Extensive knowledge, 5. Complete knowledge; and in Q.5.b) and Q.7.b) 1. None, 2. Little, 3. Average, 4. A lot, 5. Essential) ALL QUESTIONS ENDING WITH THE OPEN QUESTION ABOUT ADDITIONAL COMMENTS, ARGUMENTS, IDEAS OR EXPLANATIONS ABOUT THEIR ANSWERS Q.1. What do you know about the draft that the IASB has issued concerning its proposals for a type of disclosure labelled: “Management Commentary”? Q.2 Do you think MC increases usefulness of financial reporting for investment decisions. Q.3.a) Do you agree with the IASB decision to develop a guidance document for the preparation and presentation of management commentary instead of a compulsory IFRS? Q.3.b). If finally we have a non-mandatory guide at present, do you consider it would be convenient to have a mandatory standard in the next future? Q.4. The ED (par 9) considers “the needs of existing and potential capital providers, as the primary users of financial reports, are paramount when managers consider what information to include in MC”. The following potential users should be addressed by the MC. How much do you agree in each case? Q.5.a) The Conceptual framework ED (May 2008) considers the fundamental qualitative characteristics of general purpose financial reports are relevance and faithful representation and the enhancing qualitative characteristics are comparability (over time and between entities) verifiability, timeliness and understandability. The ED considers “the desirable characteristics of MC are those set out in the May 2008 framework ED”. Do you agree that those characteristics are desirable for MC purpose? Q.5.b) Indicate the importance which you attribute to each of the named characteristics in relation to the MC Q.5.c) In order to achieve the characteristics of understandability, the Board should recommend the use of plain language, because very technical and difficult disclosure could be against the aim and usefulness of the document. Dou you agree? Q.6. The IASB has decided to adopt a high-level principles-based approach to MC. Do you agree with this? Q.7.a).In spite of the adoption of a principle-based approach, the ED presents a list of content elements. Do you agree with the offering of a list of content elements? Q.7.b) Considering the content elements described in paragraphs 26-35 (listed in the questionnaire) value them considering their importance. Q.8 The Board’s decision is not to include detailed application guidance neither illustrative examples in the final management commentary guidance document. Do you agree with Board’s decision? Q.9 The IASB has decided to defer development of placement principles (where to disclose MC vs. Notes) until phase E of Conceptual Framework is completed (in some years). Do you agree with Board’s decision? Q.10 Is there a compulsory MC in your country’s legislation? How much time have you needed to finish the questionnaire? Both quantitative (distribution statistics) and qualitative (arguments and comments of the experts) have been considered for the next round. The second questionnaire has basically the same questions (in some cases with some minor formal modifications) but adding arguments and feedback obtained from the panel’s answers to the first questionnaire. The objective of this second round is to get a higher level of consensus giving the opportunity to the experts to change their views and/or adding comments or arguments considering the other experts responses (which are always anonymous during the process) 4.2. The results It can be concluded that there is already high consensus for several answers but some controversial aspects are detected in relation with: (i) The consideration of a mandatory versus non-mandatory guide or standard. (ii) The qualitative characteristics specially comparability (between companies), verifiability and neutrality. (iii) Application guides and illustrative examples The literature establishes that the responses may be different if the panellists had different characteristics (academics, users, auditors, preparers...). In order to compare this characteristic of the panel, a hierarchical cluster was the first analysis we performed. This procedure tries to identify relatively uniform groups of cases (agents) based on the selected characteristics. From the dendogram2 it can be concluded that, in general, there is not a clear difference between groups so, excluding some specific questions as we will comment later, we cannot say there is a significant difference in the answers as a whole depending on the background of the responder so we cannot make groups with “similar characteristics” After receiving the responses to the second round we can conclude we have not get consensus in the controversial matters but we have got stability in all the answers from the statistical point of view with the exception of two questions for which the change in the answers from the first to the second round can be considered significant3. Anyway 2 3 From number 1 to 11 are academics, from 12 to 16 and 22 listed companies, 17 and 18 are investing companies In our case the dates are qualitative, and the samples are related and quite small, so we chose the test of the signs from other no parametric methods. The test of the signs is used to resist hypothesis on the centralization parameter, and is used we think that a third round wouldn´t add any significant input and wouldn´t allow to get a higher consensus so we consider the process must stop with this second round. What we really get in this round is more feedback and arguments from the experts to explain their positions The statistical results for each question in both rounds are shown in table 2 and the results of the stability test are shown in table 3 were it can be observed that stability has not been get for questions question Q.7.b c) (in relation with resources, risks and relationships). Meaning that the changes in views about this matters have been significant in being more concrete in the requirements. fundamentally in the comparison of related samples, as in our case. The test of the signs calculates the differences between two variables for all the cases and classifies the differences like positive or negative. If the two variables have a similar distribution, the number of positive and negative differences does not differ significantly. In out work we define the level of meaning < 0,05 Table 2. Statistical results in the two rounds 1 Round N Question 1 (1 round) Question 2 (1 round) Question 3-a (1 round) Question 3-b (1 round) Question 4-a (1 round) Question 4-b (1 round) Question 4-c (1 round) Question 4-d (1 round) Question 4-e (1 round) Question 5a(1 round) Question 5b(a)(1 round) 22 Mean Mode 3,68 3 Standart Desviation ,839 Percentile 2 Round 50 (Median) 75 Min. Max 25 2 5 3,00 4,00 4,00 Question 1 (2 round) 3 5 4,00 4,00 5,00 Question 2 (2 round) 1 5 4,00 4,00 5,00 Question 3-a (2 round) 1 5 1,00 3,00 4,00 Question 3-b (2 round) 1 5 4,00 5,00 5,00 Question 4-a (2 round) 2 5 4,00 4,00 5,00 Question 4-b (2 round) 1 5 3,00 3,00 4,00 Question 4-c (2 round) 1 5 2,00 3,00 4,00 Question 4-d (2 round) 1 5 2,00 3,50 4,00 Question 4-e (2 round) 2 5 3,00 4,00 5,00 Question 5a(2 round) 1 5 4,00 5,00 5,00 Question 5b(a)(2 round) 22 4,27 4 ,550 22 4,05 4 1,046 22 2,86 1 1,490 22 4,55 5 ,912 22 4,36 4 ,727 22 3,32 3a ,995 22 2,82 3 1,220 22 3,18 4 1,296 22 3,86 4 ,990 22 4,55 5 ,912 Question 5b(b)(1 round) 22 4,41 5 ,854 2 5 4,00 Question 5b(c)(1 round) 22 4,41 5 ,734 2 5 Question 5b(d)(1 round) 22 3,27 4 1,316 1 Question 5b(e)(1 round) 22 3,50 3 1,058 Question 5b(f)(1 round) 22 4,05 4 Question 5b(g)(1 round) 22 4,18 Question 5b(h)(1 round) 22 Question 5c(1 round) Question 6(1 round) Question 7a(1 round) Question 7-b (a)(1 round) N Mean 22 3,77 3a Standart Desviation ,752 Mode Percentile 50 (Median) 75 Min. Max 25 3 5 3,00 4,00 4,00 22 4,27 4 ,550 3 5 4,00 4,00 5,00 22 4,27 5 ,827 2 5 4,00 4,00 5,00 22 3,05 3 1,463 1 5 1,75 3,00 4,25 22 4,82 5 ,501 3 5 5,00 5,00 5,00 22 4,64 5 ,581 3 5 4,00 5,00 5,00 22 3,36 3 ,727 2 5 3,00 3,00 4,00 22 2,73 3 ,935 1 5 2,00 3,00 3,00 22 3,18 4 1,140 1 5 2,00 3,00 4,00 22 3,95 4 ,950 2 5 3,00 4,00 5,00 22 4,55 5 ,963 1 5 4,00 5,00 5,00 5,00 5,00 Question 5- 22 b(b)(2 round) 4,59 5 ,666 3 5 4,00 5,00 5,00 4,00 4,50 5,00 Question 5- 22 b(c)(2 round) 4,27 4 ,767 2 5 4,00 4,00 5,00 5 2,00 3,50 4,00 Question 5- 22 b(d)(2 round) 2,95 3 ,950 1 4 2,00 3,00 4,00 1 5 3,00 3,50 4,00 Question 5- 22 b(e)(2 round) 3,32 3 ,716 2 5 3,00 3,00 4,00 ,785 3 5 3,00 4,00 5,00 Question 5- 22 b(f)(2 round) 4,14 4 ,710 3 5 4,00 4,00 5,00 5 ,958 2 5 3,00 4,50 5,00 Question 5- 22 b(g)(2 round) 4,45 5 ,858 2 5 4,00 5,00 5,00 3,32 4 1,359 1 5 2,00 3,50 4,25 Question 5- 22 b(h)(2 round) 3,32 3 1,041 1 5 3,00 3,00 4,00 22 3,50 4 1,058 2 5 2,00 4,00 22 3,59 4 ,959 2 5 3,00 4,00 4,00 22 4,14 4 ,774 2 5 4,00 4,00 22 4,27 4 ,631 3 5 4,00 4,00 5,00 22 4,09 4 ,610 3 5 4,00 4,00 22 4,09 4 ,526 3 5 4,00 4,00 4,00 22 4,36 5 ,727 3 5 4,00 4,50 4,00 Question 5c(2 round) 5,00 Question 6(2 round) 4,25 Question 7a(2 round) 5,00 Question 7-b (a)(2 round) 22 4,59 5 ,666 3 5 4,00 5,00 5,00 Question 7-b (b)(1 round) 22 4,59 5 ,666 3 5 4,00 5,00 5,00 Question 7-b 22 (b)(2 round) 4,68 5 ,646 3 5 4,75 5,00 5,00 Question 7-b (c)(1 round) 22 4,55 5 ,671 3 5 4,00 5,00 5,00 Question 7-b 22 (c)(2 round) 4,50 5 ,673 3 5 4,00 5,00 5,00 Question 7-b (d)(1 round) 22 4,36 5 1,049 1 5 4,00 5,00 5,00 Question 7-b 22 (d)(2 round) 4,45 5 1,011 1 5 4,00 5,00 5,00 Question 7-b (e)(1 round) 22 4,41 5 ,796 3 5 4,00 5,00 5,00 Question 7-b 22 (e)(2 round) 4,55 5 ,739 3 5 4,00 5,00 5,00 Question 8)(1 round) Question 9)(1 round) 22 3,45 4 1,262 1 5 2,00 4,00 3,55 4 1,143 2 5 2,00 4,00 4,25 22 3,77 4 ,922 1 5 3,00 4,00 4,25 Question 8)(2 22 round) 4,00 Question 9)(2 22 round) 3,59 4 ,908 1 5 3,00 4,00 4,00 Table 3: Stability test Question 1 Sig. exacta (bilateral) ,625 Question 2 Sig. asintót. (bilateral) ,125 a 1,000 a ,375 a ,125 a 1,000 a ,453 a 1,000 Question 5-a a ,500 Question 5b(a) a 1,000 Question 5b(b) a ,250 Question 5b(c) a ,625 1,000 Question 5b(f) a ,625 Question 5b(g) a ,219 Question 5b(h) a 1,000 Question 5-c Question 6 a 1,000 a ,500 Question 7-a Question 7-b Question 7-b (a) (b) a 1,000 a ,063 a ,500 Question 7-b (d) a ,500 Question 7-b Question 8) (e) a ,500 a ,500 Question 9) a Sig. asintót. (bilateral) Pregunta sin llegar a la consistencia 7-b -c Annex . THE RESEARCH METHODOLOGY: A DELPHI MODEL 1.1. Definition and general characteristics of the model For several decades, organizations have tried to capture the collective knowledge and experience of experts in a give field to improve decision making (Gupta and Clarke, 1996). Although the first Delphi experiment was performed in 1948 the method became popular with Dalkey and Hemer (1963) who conducted a number of Delphi experiments to reduce negative effects of group interactions in decision making. According to the classical definition, the Delphi method4 is a general way of structuring the group communication process and making it effective enough to allow a group of individuals, functioning as a whole, to deal with complex problems (Linstone and Turoff, 1975, 2002). It is a systematic process, which attempts to obtain group consensus (MacCarthy and Attirawong, 2003) resulting in much more open and in depth research than a traditional survey. Castells (1999) claims that one of the bases of the Delphi techniques is rooted in the fact that they are more socially representative than statistics based on opinions of experts in the field under investigation. The individuals involved in the group are experts in the topic. The structure of the group is designed by a monitor or monitor team that formulates a reiterative survey to address the research topic. The survey is sent to the designated group of experts (known as the Delphi panel) who then anonymously rank their preferences regarding a continuum of answers related to a series of questions or propositions posed. The experts subsequently return their responses to the Delphi monitor. Each of the iterative mail-outs of the survey is called a round, and rounds continue until stable responses between rounds are achieved. A key advantage of the approach is that it avoids direct confrontation of the experts. (Sánchez et al, 1999). This allows richer argumentation and better validation of arguments. It has the advantages of research techniques based on group interaction, but 4 Question 5b(d) a 1,000 Question 5b(e) Sig. exacta (bilateral) Question 3-a Question 3-b Question 4-a Question 4-b Question 4-c Question 4-d Question 4-e a A complete analysis and explanation of the delphi method and extensive literature review can be found in Gupta and Clarke (1996) and in Landeta ( ….) ,687a ,125a ensuring that all answers are anonymous and, therefore feedback is unbiased as to the source of the original opinion. So, it documents the opinion of the panellists, while avoiding the pitfalls of face-to-face interaction such as group conflict and individual dominance. As B point out, direct confrontation “often induces the hasty formulation of preconceived notions, an inclination to close one´s mind to novel ideas, a tendency to defend a stand once taken, or, alternatively and sometimes alternately, a predisposition to be swayed by persuasively stated opinions of others”. At the same time controlled feedback and anonymity can help panellist to review their views without publicly admitting that they have done so, encouraging them to take up a more personal viewpoint rather than an institutional position Although achieving consensus seems to be the main goal of the Delphi method, Delphi literature suggests that consensus criteria should be applied only after response stability has been established (see Chaffin and Talley, 1980; Dajani et al, 1979; Regier, 1986; Sharma and Gupta, 1993). As Scheibe et al (1975, page 262) observe, using response stability rather than consensus levels as the stopping criterion “allows much more information to be derived from the Delphi”. Stability is a measure of the extent and degree to which panel members are selecting the same responses between successive rounds. In fact, the possible outputs that could be got would be: complete consensus, majority, bimodality, bipolarity, plurality or disagreement (see Novakowski and Wellar, 2008 ) As Nielsen and Thangadurai (2007) point out “the Delphi method is based on a dialectical inquiry that encourages the sharing and exploring of divergent point of view. The emphasis is not to secure a single, universal truth, but on the range of quality ideas it generates, not only those around which consensus may form, since they may be less important to current investigations” “now a useful product of the Delphi method is crystallization of reasons for dis-sensus” (Gordon, 2004) Although there are a number of different types of Delphi exercises, they can usually be assigned to one of three broad categories (Novakowski, and Wellar, 2008) 1. Normative or regulatory Delphi. Obtaining a consensus about a preferred future 2. Forecasting Delphi. Making future predictions 3. Policy Delphi. Exploring a question of interest or with political consequences The majority of de Delphi efforts during the first decade were for pure forecasting Being the use of the Delphi method, the majority can be classified within one of the following categories. (Novakowski and Wellar, 2008): 4. Regulatory Delphi: 5. Forecasting Delphi: Political Delphi: Exploring a question of interest or with the normative Delphi the one that deals with “exercises are explorations of what should be” and so the type which is more suitable for the aim of this study. Even though there is a great diversity in political consequences As we have already established, the purpose of the study and the dimension of the application of the theory are regulatory; therefore, our work falls within the first Delphi group. In these sense there are not many studies relating with financial information using this methodology, being remarkable for example Meritum...... about the disclosure of information related with intangible assets, or Cañibano and Alberto (2008) in relation with the “enforcement” of IFRS. Related with the normalizing process coincides with the current adoption by the EU of the IFRS, Julia and Polo (2006) tried to develop an adjustment of the accounting standards to the co-operative societies using Delphi technique. We therefore have opted to use the structure developed by Novakowski and Wellar (2008) in the use of the regulatory Delphi, who establish the following steps for its correct development: Step 1: Reviewing the literature. Step 2: Preparing the questionnaire. 1.2. The development of the research According to Loo (2003) particular attention must be paid to the detailed planning and then effective and execution. (1) (2) (3) (4) Problem definition; panel selection; determining the panel size; and conducting the Delphi rounds 1.2.1. The panel of experts Reid (1988) points out that one of the keys to success in this type research is appropriate selection of panel members: they should be selected for their capabilities, knowledge and independence. One important point to recognise is that a Delphi panel is purposively rather than randomly selected. The panel is therefore not representative of any target population (Goodman, 1987) The issue of respondent bias can be overcome in this study by selecting experts from diverse backgrounds with a range of interests; so all viewpoints could be expressed (Adler and Ziglio, 1996) It is interesting to divide the experts into panels (academics, practitioners, government officials...). Their size and constitution depends on the nature of the research question There is no agreement in the literature as far as the optimum number of the panel of experts is concerned and even though there have been various attempts at establishing it in a scientific manner, the conclusions have not been significant (Galanc and Mikus, 1986). Despite this lack of clear criteria, the majority of the opinions collected in the literature consider that the number must be the groups with a minimum of seven and a maximum of 30 members (MacCarthy and Atthirawong, 2003; Denzin and Lincoln, 1994; Linstone and Turoff, 1975; Landeta, 1999). More specifically, the literature advises a number between 8 and 12 experts when using the Delphi method for regulatory purposes. Novakowski and Wellar (2008) and Richey et al (1985) suggest a similar number of experts, as these authors believe that a small panel would be sufficient when developing appropriate opinions in a consensus manner. It is important t remark that Delphi study does not aim to be a statistically representative sample of a study target population, but rather it is a group decision mechanism, which requires qualified experts with an in-depth knowledge of the study target (Okoli and Pawloski, 2004). This paper establishes two of experts recognised by the literature (Linstone and Turoff, 1975; Landeta, 1999): experts in the area (academics) and agents involved (preparers, auditors and users). This structure permits potential comparison of the different stakeholder groups, and could allow to obtain a sufficient number of perspectives from the “inside” and to perform analyses to see if there are differences in perspectives, which can be also interesting. It is essential that the experts selected receive information about the objectives of the study, the estimated time required for their participation, the potential of the research and possible benefits they can obtain by participating in it, and it is important to remark that the participants must not know the identities of the other members of the group when expressing their opinions. 1.2.2. The questionnaires Another basic aspect of the successful use of this methodology is rooted in the writing of the questions to be included in the different questionnaires. They must be clear and concise, and correctly understood by the experts. In the first phase, it is advisable to begin with open-ended questions (Heras, et al, 2006) The cover letter is especially important in a Delphi study because members must be informed and motivated about participating in all rounds and in returning their completed questionnaires in a timely manner so that analyses can be conducted and both the feedback report and next questionnaire can be constructed and distributed for successive rounds. Adler and Ziglio (1996) considers that the emphasis of the first questionnaire round (R1 is to allow the respondents freedom to define the topics that should be included in the analysis, thus minimising the degree of influence from the monitor group. Respondents are given the opportunity to indicate where they felt the questionnaire had not sufficiently addressed issues. Okoli and Pawlowski (2008) consider that no single questionnaire should take more than 30 min. to complete After the first round, once the responses have been received, the work with the results begins. Both quantitative and qualitative analyses are performed on the returned questionnaires to prepare a feedback report for the panel, as well as to assist the moderator in preparing materials for the next round. So, the monitor reviews and summarizes the responses, and then employs a measure of central tendency (usually the mean, median, or mode) to indicate where the majority of the panel responses are located on the response continuum. In the questions permitting it, the interquartile range of the responses is also estimated, as a measurement of its dispersion (Landeta, 1999). The response continuum can be based on Likert scale categories of five to seven points, or on some other rationale, in order to indicate the degree to which panel members agree or disagree with the questions or propositions posed (Critcher and Gladstone, 1998). The monitor then develops a second-round survey that reveals the response dispersion of the panel, with the different statistical indicators, such as the mean, the mode, the typical deviation and the percentiles and also includes the feedback obtained from the first round’s open-ended question(s) to ensure that the survey author has not overlooked something relevant to the topic. Questionnaire items in the second and following rounds might become more specific or precise, focussing on areas where consensus has not yet been achieved. Upon receipt of the second round, the experts are requested to reconsider their first estimations, if they consider it necessary, giving to the experts an exact copy of their responses to the first questionnaire. The process is repeated until the responds stabilise, that is, when the median shows practically no oscillation and the interquartile space stops getting narrower. This indicates that, following and anonymous exchange of information, maximum consensus has been reached (MacCarthy and Atthirawong, 2003) Although the process of response and reiteration can be repeated as many times as required, Delphi practice has revealed that the rate of response convergence is highest between rounds 1 and 2 (Linstone and Turoff, 1975a). Following the final round, the moderator prepares a comprehensive report and distributes it or a short version to all members 1.2.3. Advantages and limitations of the Delphi method The Delphi method, like all methodologies, has attracted its share of critics. Fortunately, a large body of literature has accumulated to demonstrate the usefulness of the method when well-designed and executed (Loo, 2003) Some researchers (trained in positivism with its emphasis on large sample sizes, and preferably random samples) might find these sample sizes small relative to the hundreds required for surveys, given considerations or measurement error, respondent bias, and the need for statistical power. However, the careful selection of the experts is a key factor in the Delphi method that enables a researcher confidently to use a small panel. According to Okoli and Pawlowski (2004) the Delphi group size does not depend on statistical power, but rather on group dynamics for arriving at consensus among experts. Following Nielsen and Thangadurai (2007) statistical information matters less than knowledge about the behaviours, opinions, attitudes and aspirations of people. Reliability and validity are critical properties of measures in all types of research in which questionnaires are involved. Delphi method has specific critics, claiming that the reliability of measures obtained from judgments is questionable. Hasson et al. (2000) define reliability as the extent to which a procedure produces similar results under different conditions at all times. Given that responses from different panels to the same questions con differ substantially, that the consensus achieved in later rounds might be due more to some pressure to conform than to a genuine converging consensus of opinions, and that the sue of open-ended questions con make it difficult to assess measurement reliability and validity. On the other hand Keeney et al. (2001) found a great deal of criticism accusing the Delphi technique as having no test of reliability. There is also a problem with validating numerical values attributed to expert opinion. In defence of this method from such criticisms when conducting a policy Delphi, it can be said, according to Loo (2003), that one should not necessarily expect to achieve consensus or a decision. Rather, two or more potentially conflicting policy directions might emerge, and such a result would not necessarily mean that the study lacks reliability or not valid. In fact, such an outcome might be desirable to provide the client (e.g. policy marker) with options and to stimulate the client to critically evaluate options in the decision-making process. Delphi method deserves serious consideration because the careful design and execution of a Delphi study should lead to useful findings for policy makers and program managers. Fortunately, there is a large, long-established, open literature on designing Delphi studies and examples of successful Delphis (e.g. Bijl, 1992; Loo, 1996, 1997). For an excellent discussion of the advantages and disadvantages of the Delphi method refer to Rowe et al (1991), Wounderberg (1991) and Okoli and Pawlowski (2004) Much remains to be discussed, but that discussion would obviously extend beyond the objectives of this article. We will limit ourselves to pointing out, as a means of reflection, that Delphi technique as applied here can be effective in identifying areas of agreement, disagreement and where information is lacking and where further research is required. Therefore although there are limitations to the Delphi methodology, it is concluded in the literature that the findings provide useful guidance for further consideration of the complex issue. As Nakatsu and Iakovou , 2009 point out, the Delphi method provides a good solution that allowed us to conduct an investigation with rigor and internal consistency, while allowing to produce results efficiently and with external validity. Bibliography Adler, Michael, and Erio Ziglio (eds.) (1996). Gazing into the Oracle: The Delphi Method and its Application to Social Policy and Public Health . London: Jessica Kingsley Publishers. Bijl, R. (1992), "Delphi in a future scenario study on mental health and mental health care", Future, Vol. 2 No.3, pp.232-50. Cañibano L. and Alberto f. (2008):”El control institucional de la información financiera: aplicación de un estudio DELPHI”. Vol. XXXVII Núm. 140 Octubre-Diciembre 795-829 Castells, M. (1999), “La era de la información. Economía, sociedad y cultura, Vol. 1. “La sociedad red”, Alianza Editorial, Madrid Chaffin W. and Talley W, (1980):” Individual stability in Delphi studies” Technological Forecasting and Social Change, Vol. 16, 1, January, Pages 67-73 Critcher C and Gladstone B, (1998) ``Utilizing the Delphi technique in policy discussion: a case study of a privatized utility in Britain'' Public Administration , 76, pp. 431- 449 Dajani, J. S., Sincoff, M. Z., & Talley, W. K. (1979). Stability and agreement criteria for the termination of Delphi studies. Technological Forecasting and Social Change, 13, 83-90. Dalkey N.C. and Helmer, O. (1964):”An experimental application of the delphi method to use of experts”, Management Science, Vol. 9, pp 458-467. Denzin, N.; Lincoln, Y. (1994): Handbook of qualitative research, Sage. Thousand Oaks. Galanc, T. and Mikus, J. (1986): ”The choise of an optimum group of experts”, Techinal Forescating and Social Change, Vol. 30, pp 245-250 Goodman, C. M. (1987). The Delphi technique: A critique. Journal of Advanced Nursing, 12, pp. 729734. Gordon, T.J., (2004). The Delphi Method. In: Glenn, J.C., Gordon, T.J. (Eds.), Futures Research Methodology. American Council for The United Nations University Millennium Project, Washington, D.C. Gupta U.G. and Clarke (1996):”Theory and aplications of the Delphi technique: A bibliography (1975-1994)”, Technological Forecasting and Social Change, Volume 53, Number 2, October, pp. 185-211 Hasson, F., Keeney, S. and McKenna, H., 2000. Research guidelines for the Delphi survey technique. Journal of Advanced Nursing 32, 1008–1015 Heras I, Arana G. and Casadesús M (2006): “A Delphi study on motivation for ISO 9000 and EFQM”; International Journal of Quality & Reliability Management, Vol. 23 No. 7, pp. 807-827 Julia J.F. and Polo F. (2006) La adaptación de las normas contables a las sociedades cooperativas con especial referencia a los fondos propios. Una aplicación del método Delphi. Revista Española de Financiación y Contabilidad Vol. XXXV Núm. 132 Enero-Marzo 789-816 Keeney, S., Hasson, F., McKenna, H.P., 2001. A critical review of the Delphi technique as a research methodology for nursing. International Journal of Nursing Studies Vol. 38, pp. 195–200. Landeta, J. (1999): “El método Delphi : una técnica de previsión para la incertidumbre”, Editorial Ariel, Barcelona. Landeta, J. (2006):”Current validity of the Delphi method in social sciences”, Technological Forecasting & Social Change, 73, 467-482. Linstone H, Turoff M, 1975a, ``Introduction'', in The Delphi Method: Techniques and Applications Eds H Linstone, M Turoff (Addison-Wesley, Don Mills, ON) pp 3-12 Linstone H, Turoff M, 1975b, ``Evaluation: introduction'', in The Delphi Method: Techniques and Applications Eds H Linstone, M Turoff (Addison-Wesley, Don Mills, ON) pp 229 - 235 Linstone, H.A. and Turoff, M. (1975): “The Delphi Method: Techniques and Applications”, AddisonWesley, London. Linstone, H.A. and Turoff, M.(2002): “The Delphi Method: Techniques and Applications”, at www.is.njit.edu/pubs/delphibook Loo, R and Thorpe K. (2003):” A Delphi study forecasting management training and development for first-line nurse managers” Journal of Management Development, Vol. 22 No. 9, pp. 824834 Loo, R. (1996), “Managing workplace stress: a Canadian Delphi study among human resource managers”, Work and Stress, Vol. 10 No. 2, pp. 183-9. Loo, R. (1997), “The future of management training in Canadian healthcare organizations”, Journal of Management Development, Vol. 16 No. 9, pp. 680-9. MacCarthy, B.L. and Atthirawong, W. (2003), “Factors affecting location decisions in international operations: a Delphi study”, International Journal of Operations & Production Management, Vol. 23 No. 7, p. 794. Martino, J. (1999): “Thirty years of change and stability”, Technological Forescasting and Social Change, Vol. 62, pp 13-18 Nakatsu R. and Iakovou C., (2009):” A comparative study of important risk factors involved in offshore and domestic outsourcing of software development projects: A two-panel Delphi study” Information & Management, Vol. 46, pp. 57–68 Nielsen C. and Thangadurai M. (2007):” Janus and the Delphi Oracle: Entering the new world of international business research” Journal of International Management, Vo. 13, 2, June, pp 147-163 Novakowski, N. and Wellar, B. (2008): “Using the Delphi technique in normative planning research: methodological design considerations”, Environment and Planning A, Vol. 40, 1485-1500 Okoli, C. and Pawlowski, S. (2004): “The Delphi method as a research tool: an example, design considerations and applications”, Information & Management, Vol. 42, Nº 1, 15-29. Regier W, (1986): ``Directions in Delphi developments: dissertations and their quality'' Technological Forecasting and Social Change, Vol. 29, pp. 195-204 Reid, N. (1988), "The Delphi technique: its contribution to the evaluation of professional practice", in Ellis, R. (Eds),Professional Competence and Quality Assurance in the Caring Professions, Chapman Hall, London, . Richey, J.; Mar, B. and Horner, R. (1985): ”The Delphi tecnique in enviromental assessment: implementation and effectiveness”, Journal of Enviromental Management, Vol 21, pp 135146 Rowe, G. and Wright, G. (1999): “The Delphi technique as a forecasting tool: issues and analysis”, International Journal of Forecasting, Vol. 15, nº 4, 353-375. Sanchez P.; Caminade C. and Escobar C. (1999):” En busca de una teoria sobre la medicion y gestión de los intangibles en la empresa: Una proximación metodológica” Revista VAsca de Economía Ekonomiaz, nº 45, pp. 188-213. Scheibe M, and SkutschM, Scofer J, 1975,``Experiments in Delphi methodology'', in The Delphi Method: Techniques and Applications Eds H Linstone, M Turoff (Addison-Wesley, Don Mills, ON) pp 262- 287 Sharma H, and Gupta A, 1993, ``Present and future status of system waste: a national-level Delphi in India'' Technological Forecasting and Social Change Vol. 44, pp. 199-218 Ziglio E, 1996, ``The Delphi method and its contribution to decision-making'', in Gazing into the Oracle:The Delphi Method and its Application to Social Policy and Public Health Eds M Adler, E Ziglio (Jessica Kingsley, London) pp 3- 33