The number of different violations reported was significantly greater in those who reported that the violations had taken place at police stations.70 Thus, different patterns of abuse and perpetrators emerged from this analysis, but these would largely be expected. Elections As with the public domain data, an analysis was carried to examine the relationship between violations and elections. A majority of the reports came in the periods outside of elections; i.e., in 2001, 2003, 2004, and 2006. A number of differences were found in the geographical distribution of the alleged violations, and these were somewhat unexpected. The only association with elections was in Manicaland Province, and significant differences were found in Mashonaland West, Masvingo, and the Midlands, where the violations were not associated with elections.71 There were no major differences in the alleged perpetrators, and the only difference found was that the ZRP Riot Squad were more likely to have been involved during election times.72 Violations were more likely to occur at a police station at non-election times.73 The MDC were targeted during elections and at other times, whilst organisations like the NCA, the Combined Harare Residents’ Association [CHRA], and the ZCTU were targets outside of elections.74 Property destruction and displacements were associated with elections, as has been frequently noted before, whilst death threats were more common outside of elections. Otherwise, all the other violations were as likely during elections as at other times.75 Thus, it does not appear from cases involving the police that elections result in more frequent violations by the police; rather that these alleged violations have been a continuous feature of the past years. Severity of ill-treatment: Multiple versus Single violations 771 [63%] cases reported experiencing more than one violation type, which gives an indication of the severity of ill-treatment. The data did not indicate whether a person had experienced 70 Number of different violations: Police station v no police station [2.36 v 1.88; p=0.00001]. 71 See Table 12, Appendix 3. 72 See Table 13, Appendix 3. 73 This was significantly different on a test of frequency: Violations at police station [non-election versus election; 22% v 13%; p0.005]. 74 See Table 14, Appendix 3. 75 See Table 15, Appendix 3. Who guards the guards? Violations by the Zimbabwe Republic Police, 2000 to 2006