ActionAid International - Combined Harare Residents’ Association - A Study on the impact of “Operation Murambatsvina/Restore Order” researchers would allocate each other specific areas of the ward for the two days of data collection. Therefore, a representative coverage of the wards was met. 2.2 Sample size Initially, CHRA wanted to visit each and every homestead to establish the true extend to which the “Operation Murambatsvina” had affected households. However, due to the lack of manpower and resources for this exercise only a third of the homesteads were visited in the affected wards. This is a statistically significant proportion of the communities to make sound inferences about the general population. Consequently, at least 500 homesteads per ward were visited during the exercise. 2.3 Limitations of the survey The Operation Restore Order survey has several limitations which must be considered when interpreting and using the results to judge the impact of the Operation. First and foremost, the time frame covering the survey is approximately 2 months after the first documented demolitions. Many changes may take place between the survey data collection, entry and report write-up. Thus there may be real changes taking place that will not be revealed by the household survey alone. For this reason it is important to also consider contextual information as well as qualitative information collected. Additional limitations are captured below were noted by the team: ✓ Survey coverage: Initially, CHRA wanted to collect information possible from all residents in Harare of the affected wards, however, fuel, money, researcher safety and manpower were limiting. The survey therefore only covered the high density housing areas of the city. The Survey did not include Chitungwiza and this remains an area that should be assessed. ✓ Comprehensive information: The information collected in the questionnaire is not comprehensive, due to the debate between coverage (numbers) and in-depth interview (detailed analysis), a compromise was therefore the one third sample. Additional information such as ART, HBC and mobility and migration were all not captured. ✓ Coverage of holding camps: This survey did not take place in the holding camps and informal settlements. ✓ Delayed implementation of the survey: The first demolitions started on 25 May 2005, and this report is coming out two months after. ✓ The time frame: Like all studies conducted over a short period of time the final product is a mere snapshot of the bigger picture. While every effort has been made to paint as accurate a picture as possible of the household impact of the Operation, the representative sample used in this regard is 33.3% of the total number of homesteads in the 26 wards. ✓ Baseline data and the extent of the impact: The study is conducted within the context of various views and differing understanding of impact of the Operation and the current situation. For example, government reports estimates are inconsistent with those of the United Nations - IOM. In the absence of reliable and consistent baseline data the evaluators were dependent on using a sampling method in capturing as much information as possible and applying this to an agreed understanding of the factors that constitute impact of the Operation. ✓ Additional Variables: The study did not capture access to HIV/AIDS quality care and treatment (ART and HBC). The mobility and migratory patterns of affected communities were not captured. The study did not interrogate the cost 4