Reliability and Validity of Qualitative Research

Question: Discuss about the Reliability and Validity of Qualitative Research. Answer: Introduction: The census data in the excel sheet was organized and chosen as per the instructions therein. Since no particular formula or trick was employed in choosing the sample, we call this method of data sampling as random sampling. Sometimes it is referred to as equal chance sampling (John, 2009). Unlike other probability sampling methods, equal chance sampling discourages biasness (Gravetter , 2011) in results hence the outcomes from a simple random sampling method are usually a representative of the whole sample. This is why it has been employed in this data selection. It is always a daunting task to decide which test statistic to apply so as to compare sets of data. It is more difficult to a non-statistician or even mathematician(Golafshani, 2003). However, statistical methods advice that the characteristics of the data to be compared should be able to guide us on which test statistic to use. First, we should ask ourselves whether the data is ordinal, ratio or nominal. Secondly, is the data paired or not paired?(Utts, 2014). The number of samples to be compared also determines the test statistics to be employed (David George , 2006). Considering all the mentioned criteria, the best test statistic that would fit this kind of data was analysis of variance (ANOVA). It is ANOVA since we are dealing with more than two samples. (Peace Parrillo, 2008). The generation of income distribution is all shown in excel document attached in the second sheet. Going by the descriptive analysis using excel, it can be confidently said that majority of the people are in the very poor category. This category compared to the rest has got the largest number of people who are earning between 110 and 914.25. It has 58 people in number. It is then followed by the poor who earn between 914.25 and 1718.5. These are 29 in number. From statistical perspective, it can be concluded that the data is not normally distributed. It is skewed towards one side and not just normally distributed on both ends of the normal curve. The distribution of the income for the families is not normally distributed because there is a large proportion of people earning less and a few people earning more. Normal distribution always has a balanced scenario. In this case the people who earn less and those who earn more should be the least in number then those who earn moderately should be fairly large. When determining the top 10% value, in a set of data, it is first arranged in order for example in ascending order. In percentiles, a given data is assumed to have been divided into 100 equal parts. So, the bottom 10% value represents the 10th percentile while the top 10% value represents the 90th The excel formula that has been employed to determine the percentiles is =PERCENTILE.EXEC (RANGE, K) k represents the fraction of the actual percentile that we need to establish. In our census data the bottom 10% value in WEF is 83.9 and the top 10% is 507.The chance that a family income is less than or equal to 200 is 0.09. References David, S., George , P. (2006). Introduction to the practice of Statistics. Freeman. Golafshani, N. (2003). Understanding reliability and validity of qualitative research. Gravetter , F. J. (2011). Research methods for behavioral sciences.