Election Investigations Guidebook knowledge of facts, opinion, belief, information or evidence related to the investigation”).173 Investigators may need to differentiate between these categories when deciding whether or not to interview a particular individual. In some cases, it may not be necessary to interview a complainant, especially if the initial complaint must already contain “a clear, precise, and detailed” account of the underlying facts, as in Costa Rica.174 On the other hand, Bhutan’s Election Dispute Resolution Manual cautions investigators to question all witnesses in order to avoid missing pertinent facts.175 In short, states should allow investigators to use their discretion to determine whether or not individuals should be interviewed during the course of an election investigation. In systems that allow interviews and when investigators are well trained and prepared to conduct and accurately record the interviews, conducting interviews can be considered a best practice because it provides the arbiter with direct testimony from relevant and critical sources. Preparing for Interviews According to the Canadian Special Investigators’ Manual, developing written questions is “[t]he most important aspect of a successful interview.”176 Prior to an interview, investigators should prepare a set of written questions to guide the conversation. A written interview plan provides a framework for questioning and also allows investigators to best assess the evidence by compelling 173 See Comm’r of Canada Elections, Ch. 7 Official Cautions, in Investigators’ Manual 1–20 (2000) [hereinafter Canadian Investigators’ Manual Ch. 7] (defining “witness” in the context of an election investigation). N.B. on terminology: Different sources use these terms in different ways. The UNDP Investigation Guidelines, for example, use “witness” as a synonym for interviewee. United Nations Dev. Programme, Office of Audit & Investigations, Investigation Guidelines ¶ 11.3 (Jul. 1, 2010) [hereinafter UNDP Investigation Guidelines 2010] (broadly defining witnesses as “individuals making the allegation; victims of the alleged act; individuals with direct or indirect knowledge of anything relevant to the investigation; individuals with good knowledge of business processes related to the alleged act; or experts”). Similarly, “subject” can refer to either an interviewee generally or to a suspect. Compare Christopher D. Hoffman, Int’l Found. for Protection Officers, Investigative Interviewing: Strategies and Techniques 2 (2005), http://www.ifpo.org/wp-content/uploads/‌2013/08/interviewing.pdf (using “subject” to mean any interviewee) with Uniform Guidelines 2009, supra note 25, ¶ 37, n. 8 (using “subject” as synonymous with “suspect”). For clarity, throughout this section, we use “interviewee” to refer to any individual who is interviewed by an election investigator. Where it is necessary to specify further, we use “witness” as defined in the text above, “complainant” to indicate an individual submitting an allegation of election violation, and “suspect” to refer to an individual suspected of committing an election violation. 174 Ley No. 8765, Código Electoral [Electoral Code], tit. V, ch. VIII, art. 267, La Gaceta, Diario Oficial [L.G.], 2 Sept. 2009; accord Bhutan Election Dispute Settlement Manual, supra note 27, ¶¶ 6.4–6.5. (“It may not be necessary to question the complainant as a witness if the facts of the case are clearly mentioned in the complaint letter. On the other hand, the complainant may be questioned in case there are some issues and facts which need to be verified or corroborated further.”). 175 Bhutan Election Dispute Settlement Manual, supra note 27, ¶ 6.9. 176 Canadian Investigators’ Manual Ch. 11, supra note 168, at 10; accord Ralph Crenshaw, Stuart Cullen, & Tom Williamson, Investigative Interviewing: Best Practice in Questioning Witnesses and Suspects, in Human Rights and Policing 254, 251–258 (2d ed., 2007) (discussing a “written interview plan” as an “important element” of the interview process). 70