**2. Methodology**

The data used in this chapter was elicited from 140 household questionnaires administered across 16 wards of Chivi District and focus group discussions held in six wards of Chivi district. This data was also supported by data from key informant interviews held with three Non-Governmental Organisations (NGOs) and 16 Agricultural Research and Extension (AREX) officials.

Data capturing was organized in Microsoft Excel 2013 and later transferred to Stastical Package for Social Science (SPSS). Prior to the analysis, captured data was coded according to the levels of measurement. This allowed for uni- and bivariate data analyses. Data analysis was done using SPSS version 22. Chi square and Cramer's V value were calculated and analysis was set at 0.05 confidence level. In order to describe and identify relationships that must be taken into account and characterise CA project in Chivi District, frequency tables and bar graphs were generated (univariate analysis). Frequency distributions described the number of times the different attributes of a variable were observed in a sample. This allowed for the comparison of different variables. Statistical tests of significance were conducted on the levels of awareness and general perceptions in order to explore independent variables e.g. gender; age; level of education differences. Chi-square tests was used to calculate significant differences in different demographic groups on their adoption and practices in the Conservation Agriculture project [14]. A 95% level of significance was used, which is most commonly used in social research [15].

Cramer's V test was used to measure the strength of relationships. It measures the strength of relationship for any size of contingency table, and it offers good norming values from zero to one (0–1) for relative comparison of the strength of correlation regardless of the table size. For Cramer's V, 0.0 to 0.30, the strength is considered no relationship to weak; for Cramer's V, 0.31 to 0.70, the strength is considered moderate relationship; while for Cramer's V from 0.71 to 1.0, the strength of the relationship is considered strong [16].

For qualitative data analysis, Archive of Technology, Life world and Language. Text interpretation (Atlas.ti 8) was used to analyse data from the household

**165**

**Figure 1.**

*CA adoption in Chivi.*

*Adoption of Conservation Agriculture as a Disaster Risk Reduction Tool in Chivi District…*

questionnaires, focus group discussions and key informant interviews. Tools such as Co oc was used for comparisons using the occurrence frequency, Co-code Doc Table for numeric analysis as well as Networks and Report tool for visual and text analysis. Results from Atlas.ti 8 were used to compliment data from SPSS. Results

The physical adoption of the CA project was measured based on project adoption

Only 30% of households in Chivi are practicing CA, refer to **Figure 1**. AREX officials and Focus Group Discussants showed that CA started as early as 1995 in some wards such as Ward 10 but became more popular from 2008 when the government of Zimbabwe formalized it. This implies that the project has been long operating in the District, despite low adoption percentage. However, CA benefits are normally realized at least after 10 years of practice [17]. Hence a 30% adoption is not that low, considering that the project is formally slightly over a decade in most wards. After seeing the benefits more farmers are likely to adopt CA. However data on CA adoption trends did not support this. Key informants confirmed a decline in adoption trend over the years in all wards. In ward 21 of the 300 farmers who initially adopted CA in 2008 only 80 are currently practicing it. Of interest is that Ward 21 was listed as the third highest adopter of CA in the District by NGOs. This gives a gloomy picture to the sustainability of CA as a drought risk reduction tool in

To get an insight into the spatial adoption of CA and the long term plans of farmers on CA, plot sizes were also assessed. Key informants showed that farmers under the main NGO, CARE increased their demo plots from the 18 mother demo plots of 1 hectare to 180 baby demo plots across its 12 wards. However the questionnaire survey showed that 100% of CA farmers are still working on demonstration plots in groups and have not adopted the full CA package onto their individual

*DOI: http://dx.doi.org/10.5772/intechopen.94318*

records of NGOs operating in Chivi district.

**3. Results and discussion**

**3.1 CA adoption in Chivi**

the District.

**3.2 Extension of CA plots**

were presented as graphs, charts, visuals and narratives.

*Adoption of Conservation Agriculture as a Disaster Risk Reduction Tool in Chivi District… DOI: http://dx.doi.org/10.5772/intechopen.94318*

questionnaires, focus group discussions and key informant interviews. Tools such as Co oc was used for comparisons using the occurrence frequency, Co-code Doc Table for numeric analysis as well as Networks and Report tool for visual and text analysis. Results from Atlas.ti 8 were used to compliment data from SPSS. Results were presented as graphs, charts, visuals and narratives.
