There are interesting clusters of relationships when correlations between the alleged perpetrators are calculated. Non-state actors – ZANU PF Youth, ZANU PF supporters, and “war veterans” – are generally reported together, but, interestingly, there is a negative relationship between ZANU PF Youth and ZANU PF supporters, but not between each of these and “war veterans”. 26 There is a clear pattern seen in the testimonies where the victims report either ZANU PF Youth or ZANU PF supporters, and infrequently both together, but also that either of these two groups were acting in concert with “war veterans”. This suggests that “war veterans” were probably in charge of these groups, and, of course, it cannot be discounted that these “war veterans” were not uniformed forces in civilian guise, as was frequently alleged in the press reports. It is also possible, given the frequency with which ZANU PF Youth are mentioned, that these were groups of youth militia deployed into the communities, as was also alleged in many press reports. There are no associations between the various branches of the ZRP and the ZNA with the non-state actor groups, and the ZNA is always reported as operating alone. Here it is worth noting that a very large percentage of the reports alleging violations by the ZNA came from Harare, and were indicated in the reports as being retaliatory attacks for alleged violence by the MDC. 27 As regards the alleged violations, the data corroborates generally the findings of the reports of the Solidarity Peace Trust and the ZPP. No murders are reported again in this report since the Human Rights Forum did not have any direct reports of murders, and hence could not confirm the reports in the press. As can be seen, there are a number of changes observed over the three months. Assaults and Assault [GBH] increase from March to April, and then drop very slightly during May. Abductions, thefts and property destruction all increase very significantly from April to May, whilst unlawful arrests and detentions drop from March to May. The total number of all violations increased significantly from March onwards, 28 and increased further during May. The changes were more dramatic from April to May. Theft and destruction of property significantly increased from March to April, 29 and there was a significant increase in the number of witnesses to the violations mentioned. 30 It is important to note that there were so many witnesses to the violence, for this adds considerable weight to the allegations, and would allow any independent investigation to seriously test the veracity of these allegations. Additionally, it also evident that there was little attempt by the perpetrators to hide their activities, lending further credence to the frequent claims by human h groups that a climate of impunity protects the perpetrators. 26 27 28 29 30 ZANU PF Youth & war veterans [0.10; p=0.01]; ZANU PF supporters & war veterans [0.17; p=0.01]; ZANU PF Youth & ZANU PF supporters [-0.25; p=0.001]. There is the most recent report of the deployment of the army in Mutare, where it is alleged that the ZNA attacked civilians, resulting in injuries to 70 citizens. March v April: Total number of violations [p=0.05]. March v April: Theft [p=0.001]; Destruction of property [p=0.001]. March v April: Number of witnesses mentioned [p=0.01].