A research report reflects the thought process of the researcher, as well as his/her knowledge about the subject and the methods used. Good research is characterized by the following attributes:

Well defined objective: the research problem must be well defined, the objective determines the methodology to be adopted and the research strategy, it defines the scope of the research, hence it is very important to write the thesis statement precisely and correctly.

Based on Sound Methodology: the logic behind choice of research methods and techniques should be clear. Sound methodology ensures system and sequence in the reporting of research findings.

Executable: executability implies the possibility of finding and collecting sufficient data and its analysis, sometimes a solution may seem very promising but may fail when actual data collection is called for. A good research is supported by real data and findings.

Originality: this is an attribute highly rewarded by the research fraternity, the greater the degree of acknowledged novelty, the greater the rewards.

Promotes further research: a good research should be based on thorough knowledge of past and present context, should present a viable solution to the defined problem and leave a hint of further possibilities in the domain investigated.

A well conducted research may often require contribution of many minds due to the complexities involved; seeking advice from expert mentors is always a good option.

Data analysis refers to the process of cleaning, inspecting, modelling, and transforming data with the sole aim of gaining useful details, which support decision making. This term has got different approaches and features which include a variety of techniques under different names in many fields like science, business, social science, etc.

Data mining is a form of data analysis which deals with knowledge discovery and modelling for the sake of predictive purposes rather than descriptive ones. Business intelligence makes use of data analysis for the aggregation of business information. In statistics, data analysis is divided into exploratory data analysis or EDA, descriptive statistics, and confirmatory data analysis or CDA. CDA is used for falsifying or confirming the already existing hypothesis, while EDA is used for finding out new features in the concerned data. There are two other varieties of data analysis namely predictive analytics and text analytics. The former makes use of structural or statistical models for classification or predictive forecasting while the latter utilizes linguistic, structural, and statistical techniques to gain as well as segregate information from unstructured data.

Data analysis is a process, which includes several phases. Data cleaning is the first stage of a data analysis, and refers to the prevention and correction of errors in the methods through which data is entered as well as stored. There are phases in data analysis like initial analysis and main analysis. The initial phase includes four questions regarding the quality of data, quality of measurements, initial transformations, and finally relevance of the study in satisfying the intentions of a particular research design. In the main data analysis, either confirmatory or exploratory approach is adopted. In the confirmatory approach, the hypothesis related to the data is tested, while in the exploratory approach, no such hypothesis is addressed.

In short, data analysis involves a series of procedures.

In the field of statistics, the term ‘meta analysis’ refers to the different methods that contrasts and combines the outcome from various studies in order to identify certain patterns in those results, the reasons for disagreement among those, or any kind of relationships. In fact, meta analysis is carried out by identifying a common measure belonging to an effect size, of which the final output is the weighted average of the same. Here, the weighting corresponds to the sample sizes included in individual studies.

Generally, meta analysis aims to take an estimate of the actual effect size against a less accurate one which has been derived from a particular study conducted under a given set of conditions and assumptions. Therefore, a meta analysis gives the summary of different studies conducted on the same subject, thus giving the reader a wide knowledge regarding the existence and size of the effect.

In some cases, meta analysis is used for systematic review. For example, meta analysis is conducted on various clinical trials of a particular medical treatment so as to gain better knowledge about the effectiveness of the same. Meta analysis forms a part of the network namely estimation statistics which mainly relies upon confidence intervals, effect sizes, precision planning for data analysis, etc. Further, meta analysis can be used instead of null hypothesis significance testing.

Moreover, meta analysis is widely used in studies related to cost-effectiveness where the outcome from several studies are combined to determine the most effective of therapies, the estimation of the amount of effectiveness when it comes to different choices, and also the estimation of cost-related outcomes in expensive therapies.

When a meta analysis is conducted, it is necessary to include a main research question in the research protocol, a summary of the intended studies, inclusion or exclusion parameters for selecting a particular trial, and methods used for trial selection.

To summaries, meta analysis is extensively used in different fields of study.