Best Practices for Energy Price Indices © Copyright 2003, CCRO. All rights reserved. 12 submitted, the designated noncommercial group should make reasonable efforts to submit the corrected data in a timely manner. To ensure a systematic approach to error resolution in the transaction data, the CCRO recommends that index developers define, upon consultation with the data providers, whether they require data providers to submit any and all error revisions to the index developers regardless of the size or nature of the errors or whether a materiality threshold should apply and, if so, what it should be. This would ensure that there is no element of subjectivity on the part of the data provider. The benefits of a zero error revision materiality would eliminate any analysis on the part of the data providers, and the burden if any would be on the index developers to evaluate the impact of all error corrections in the data sets from the providers of data. 4.1.5 Sufficient Sample Size For a given index, the sample selection should be such that sufficient and representative transaction data is used to construct the index. The index developers should specify a minimum sample size, volume sizes, and number of companies for an index to be deemed representative. 4.1.6 Treatment of Double-Counting For each index, transactions submitted from multiple sources (buyer, seller, trading platforms and/or clearinghouses) should be identified and handled in a consistent manner as defined by the index developer that is clearly articulated, documented, and communicated. 4.1.7 Treatment of Outliers For each index, outliers, as defined by the index construction methodology developed by the developer, need to be identified and eliminated in a consistent manner that is clearly articulated, documented, and communicated. 4.1.8 Index Calculation To calculate a given index value, the index developers should aggregate all admissible data for each location and calculate an average value. Each index developer is responsible for developing, testing, and documenting the methodology used. 4.1.9 Insufficient Sample Size When the sample size is insufficient but an index value is published nonetheless, a clear indication must be provided with the published index value. The index developer could choose to generate the index value based on the submitted albeit insufficient data, on data including outliers, or using any other appropriate method. This enables index publishers to consistently provide indices that market participants can utilize even when the sample size is insufficient. As part of the published description of the methodology used to derive a specific index, the index developer should clearly document what method it will apply in such instances. 4.1.10 Establish an Error Revision Process – Following Index Publication The index developer should define a process for the resubmission of data if a data provider discovers errors, including errors of omission, after the publication of the index. This process