Given Hapunda

144 different form of data analysis may be advisable to use. However, if some observations are untrustworthy or missing altogether, additional data collection may be necessary. Before any analyses are performed, it is imperative that data is carefully considered on its levels of measurement because different statistical techniques will only work with certain levels of measurement. Data collected for all research investigations can be nominal, ordinal, interval or ratio (Black, 1999). Nominal data simply distinguishes between categories (Leary, 2004; Roberts, 2008). For example, responses of “Yes, No and I don’t know” or codes like 1 or 2 for distinguishing categories male and females. The numbers in Nominal data are treated as “markers”. They cannot be added, divided or multiplied. Ordinal data represents numbers with size that is meaningful. This level is often used for variables that cannot be directly measured such as happiness, anxiety or satisfaction. Mainly, likert scales are a good example of numeric categories that provide numerical hierarchy rather than absolute measurement (Norman, 2010). The numbers at this level indicate order while Nominal indicates difference only. The last levels of measurements are grouped as representing continuous data (Interval and Ratio scales). Data on these scales is measured directly using infinite scales where increment on scales is of equal distances (Pell & Fogelman, 2002). The difference between the two scales is that the Ratio scale has an absolute zero, meaning that a score of Zero equals absence of existence. For example, zero kilogram (kg) weight means there is no weight. In comparison, the interval scale has no absolute zero, meaning a score of zero does not mean absence of existence. For example, zero degrees Celsius does not mean the absence of temperature. When deciding on which statistical techniques to apply, the first question often is about the level of measurement. In the world today, major advancements in data analysis softwares have been made. Several quantitative data analysis packages are in existence (SPSS, mini TAB, Excel, stata, R etc). The choice of the software package is determined by the project aims and objectives in connection with the applied design beside the software cost which is one of the drawbacks for accessing some of the software. The software package used for analyses in this chapter will primarily be the Social Package for Social Sciences (SPSS) and Microsoft Excel in some cases. While quantitative analysis may involve complex statistics, much of the analyses done in typical monitoring and evaluation are straightforward and easy to understand. There are two types of statistics that are involved when analysing data quantitatively in M&E: Descriptive and Inferential statistics .