Unlike the Business Process Level section of the RREC, where the implementation method is relatively defined, there are a variety of different ways that the answers to the questions in the Record Level section can be obtained. For example, the answers could be acquired through interviews of relevant agency staff, conversations with experts such as the Agency's legal staff, group decision conferences, or surveys. The RREC, in particular, the Record Level section specifies the information that needs to be collected in attempting to identify a comprehensive set of records management requirements but does not dictate the mechanisms by which the questions get asked and answered.

The method used to answer the questions outlined in the Record Level section of the RREC should be determined in much the same way a research method would be selected to answer a research question. A variety of factors need to be considered and the most cost-effective mechanism for gathering the information should used. For example, the cost of the method should be considered (e.g. interviews tend to be a much more expensive way of gathering information from people than surveys). The organizational structure and environment must also be considered. For example, the individuals within an organization who can adequately and authoritatively answer the questions must be identified. The degree of alignment or consensus among relevant staff as to how records should be managed will also affect the method used in addressing the questions in the Record Level section of the RECC. If consensus does not already exist, then a method such as a group decision conference could be used to create consensus or alignment.

Additionally, the type of question will dictate the methods that can be used to elicit information. For example, open-ended questions that require elaboration and discussion are best answered through interviews while those with an easily identified set of answers could be addressed through the implementation of a survey. Other key factors in choosing methods for gathering records management requirements include the size of the organization, the geographic distribution of staff, and the scope of the information system under development. In cases of very large organizations, it is unlikely that information can be gathered from every staff member that will be affected by the development of an information system and therefore a representative sample of different perspectives will have to be identified and appropriately defined sub-groups of users of records will have to be sampled during the information gathering activities. The Models for Action team has concluded that the implementation of the RECC will differ substantially across organizations and that there is in fact, no one ideal way for implementation. For a more comprehensive description of methods that can be used and the advantages and disadvantages of each, please refer to CTG's Making Smart IT Choices handbook.

Given that APA is a small agency of under 40 individuals, many of whom require access to the Project record, the Models for Action team decided to include all Agency staff in the survey designed to capture internal agency uses of the project record.

The following section provides a brief description of using a survey to capture information for the Record Level section of the RREC.

• The survey was designed to identify answers the following questions in the Record Level section of the RREC:
• Which other agency business processes or purposes might require access to the project record?
• For each of these business process or purposes, what components of the record need to be accessed?
• For each of these business processes or purposes, what are the most efficient/effective ways of accessing the records (i.e. indexing)?

The questions focus on internal secondary use of project records. Based on the survey results, we have identified the types of documents or components of a record that each of the individual staff members within the agency require access to as well as their preferred mode of access in terms of indexes. While additional work may have to be done to collapse the categories of documents, the Models for Action team feels confident that a comprehensive set of documents and indexing methods have been identified from Agency staff. An additional indirect benefit to surveying the entire agency was reported to the project team by agency staff. An APA staff member indicated to the project team, that by including all staff members in the survey, buy-in from agency staff on the overall project approach was increased. In other words, the process of collecting input and assessing the needs of all agency staff has enhanced overall support for the project. While this approach may not be possible in all agencies, the survey approach which is relatively inexpensive allows for an increased degree of input from staff in an organization and further fosters feelings among staff that there will be benefits to not only those directly involved in the business process being automated but to all staff requiring access to those electronic records.