Research tools for basic research

Abstract

Statistics is an integral part of any research project. It is mainly used in sample size calculation in the planning stage and data analysis at the conclusion of the study. Furthermore, the readers of a research article should also have some understanding of the statistical methods used in order to draw appropriate conclusions from the study. Though statistical tools would allow one to derive scientifically sound conclusions from research, it will not be able to compensate for flaws in the study design. Easy accessibility to statistical computer software in modern day has made the statistical calculations much easy fro the investigators. However, a basic knowledge of statistics is required in order to select the most suitable method of analysis. Sample size calculation is a method in which the investigator would be able to know the minimum number of subjects that needs to be included in order to draw definite conclusions. Such a calculation will minimise the possibility that the observed effect is due to chance and spare both the investigator and the subjects of over-recruitment. Different sample size calculation methods are used in population based descriptive studies, two group comparisons of means or rates and in correlation and regression calculations. Data analysis is done mainly at the conclusion of subject recruitment. In order to obtain valid results the data analysis method should be decided as early as the planning stage of the study. The statistical method used would depend on the study design, the type of data and what the clinically useful conclusions that can be derived from the observations. The data would let us draw conclusions from the observations and the statistical methods will describe the validity of those conclusions. In descriptive studies statistical tools are used in data presentation where confidence intervals could be calculated for means, proportions, counts and rates. Hypothesis testing is done commonly in two group comparisons and in certain instances in three or more groups. In such analysis various statistical methods could be employed based on the characteristics of the study design, the variables and the study population. More advanced statistical analysis include regression and correlation which are used too describe the relationship between two or more variables and rnultivariate analysis used in study of multifactorial associations and confounding effects.