ZIMBABWE ELECTION SUPPORT NETWORK registration by age of voters on election day, gender, province and setting (urban and rural) by making comparison with national population figures and by making a comparison with the 2013 PVR. Objectives The voters' roll audit was guided by the following objectives: • To increase trust and citizen confidence in the conduct of the 2018 elections • To increase credibility of the 2018 BVR voters' roll • To independently verify the quality of the 2018 voters' rolls. People-to-list The people-to-list test, involves selecting individuals at random and then determine if those individuals are listed in the voters' roll and if so, whether the information is correct. A representative sample is used to select individuals. The test captures people who should be on the voters' roll, but whose names are missing. The test also determines what proportion of the eligible voting population is on the voter registry. Measurement Criteria Accuracy relates to the rate of error in entering data on the individual voter, that is, name, date of birth, gender, identification number, etc. List-to-people The list-to-people test identifies the extent to which the voters' roll is bloated with the names of nonexistent or ineligible voters. The test captures people who are on the voters' roll but should not be. The test ensured that every name on the voters' roll was that of an actual person eligible to vote and that his/her information, such as address, was correct and up-todate. This method involved randomly selecting names from the voters' roll and then sending out enumerators to locate the selected individuals, confirm their existence and the accuracy of the registration information (name, gender, date of birth, identification number etc.) of the individual. Completeness relating to the proportion of eligible voters included on the voters' roll. Currency relates to whether the information on the voters' roll is updated, for example through transfers for those who have moved to a new location and removals of records of persons who are no longer eligible to be voters for example are deceased, nonresidents or multiple registration. Methodology A Voters' Roll Audit (VRA) is a systematic and scientific assessment of the quality of the voter registry. An audit is a method that is used by independent election observers around the world and has been proven reliable and accurate internationally. Verification of the accuracy of a voter registry in the form of a VRA can help to detect and deter electoral fraud, correct administrative errors and promote broad public confidence in the process on Election Day and beyond. The goal of this VRA is to increase citizen trust and public confidence in the conduct of the 2018 harmonised elections and to increase credibility of the 2018 BVR voters' roll. Sampling List-to-people A nationally representative sample of 1,200 was drawn from the voters' roll from a sample universe of 5,683,936 million registered Zimbabweans. The field tests were conducted using the 2018 PVR. The sample was stratified by province and rural/urban domains. The sample was allocated using Probability Proportionate to Size (PPS) into the 10 provinces of Zimbabwe. The total number of people in the voters' roll in each province was used to allocate (measure of size) a sample of 1,200 individuals in the ten provinces of Zimbabwe. Within each province the allocated sample was further allocated into rural and urban areas using the proportions of populations in the rural and urban areas of the province in the voters' roll. Computer Audit One method for examining the accuracy, comprehensiveness/ completeness and currency of the voter register is a computer audit. The computer audit was used to assess both the internal and external consistency of the voters' roll. The computer test analysed variables that include 13 FINAL REPORT OF THE 2018 VOTERS' ROLL AUDIT