What are important considerations in determining an appropriate data analysis approach?



What are important considerations in determining an appropriate data analysis approach?

In the evaluation of programs and policies, there are three main approaches to data analysis : descriptive (descriptive) study, correlative and causal analysis. The choice of approach to the analysis of the data depends on what specifically about the ongoing program or policy researcher wants to know. These approaches to data analysis are similar to each other, but they can be used to evaluate various aspects of the program or policy.

Descriptive Analysis

Descriptive research in the assessment carried out in order to describe how to implement the program , what the event was held, what resources have been spent, what progress has been made, etc. This type of research does not assess specific qualitative or quantitative characteristics of the program , it optimizes the data set characteristics and evaluate the program , both qualitative and quantitative side. Descriptive statistics are trying to identify the cause of this or that phenomenon, about the program and the effect, introduced in connection with this phenomenon. Types of questions that are put in the evaluation of the program, providing a good conduct research and draw conclusions.

Descriptive (descriptive) research can be both qualitative and quantitative. In this study, can be combined as a numerical and qualitative data that the study can be combined and divided, depending, on what kind of response the program is evaluated. The data obtained during a descriptive study can be grouped into tables, graphs, and interpreted in narrative form. Descriptive study provides data to give more convenient for the evaluation form. This type of research to optimize data collection and evaluation methods to help find the statistical characteristics of and relationships between data, which allows estimation of more effective ways.

Correlative Analysis

Correlation analysis is a powerful tool, enabling them to simulate the future conduct of a program or policy. However, you should very carefully interpret the results of such research. Finding correlation between two characteristics of a program or other valued variables does not allow concluding that a characteristic is the cause of another. Lack of understanding of the general statistical rules leads to the fact that in the correlation analysis of data people receive inaccurate results. However, the need for correlation analysis often occurs when, for practical or ethical reasons cannot carry out the experiment . Ongoing program or policy cannot ...