Research Methods Week 6
Image source
The meta-Analysis (Presentation)
After 2 weeks of break, We start our Research Methos Class and it surprised me at the beginning when the tutor said that we have to do our presentation. Each group had a different topic to explain, our’s was Meta-Analysis Although we have created our presentation, Actually, we were not prepared to present that. Somehow we have mange to present that and we got some good comments from the tutor so at the end everyone was happy. The following points included in the presentation.
What is it?
What kinds of questions/problems might it be useful for?
How could it be used in IT research
What are the strengths of the approach?
What are the weaknesses of the approach
What is Meta-Analysis?
Meta-analysis is an epidemiological formal, quantitative, method of the study used to consistently analyze the findings of previous research in order to draw a summary regarding the research. The research is usually based on randomized, supervised trials, though not always
Check the Video below What is a meta-analysis?
What is a meta-analysis?
What kinds of questions/problems might it be useful for?
For this, we can use the following questions to get the required data
Integrating data from various research helps to identify the effect of studies associated with a specific outcome.
Using this question we can compare different researches and try to find the required outcome that we are looking for by this our
Helps to differentiate results from various research.
We can compare the previous Researches and get an idea about it. For instance, If we do research about a newly build Sever but, there is some research already done regarding the same Sever before and we can compare two researches and get a good idea about it. Sometimes the first research has done only about the Server Hardware. Then we have to give more priority to the software sid.
Helps to identify if new research is needed to be done.
In this, we can compare the previous research details and we can come up to a point that do we have to do the whole research aging or not. if we feel that the research that has been conducted regarding the sever is not enough we can do a whole new research aging.
It can be misleading when different types of outcomes are grouped together
When we compare two or more different researches some time those result won't match to each other and it will misguide us
Produce new speculations for future investigations.
In this kind of Question, we can directly come to a point that we will do a whole new Research again. Cause any of the above things aren't matching at all
How could it be used in IT research
As we make technological developments in computational power, new database programs have made the process even easier there are lot of databases Software that we can use to get the requires results such as Revision Manager (RevMan), Metafor (R Package), JASP these are used for meta-regression and they are user-friendly
What are the strengths of the approach?
Meta-analysis is one of the best ways of making work complex. Within a given time, a single research team can comparatively send out many results. But theoretically, it gives access to more data than any team could generate.
Inconsistency of result across studies can be quantified analyzed and corrected
Moderators can be considerate to explain variations between studies
The existence of publication bias can be investigated
What are the weaknesses of the approach
The meta-analysis, researcher must be conscious before believe on the data and statistics it provides, also has drawbacks. The main issue is the weakness for bias in publishing and twisted results. Work producing results which do not reject the null possibility that continues to remain unpublished, or risk not entering a database. Unless the study is limited to positive-resulting analysis, then the the validity of the whole project is undermined.
Discourage large unconditional trails.
Raises the trend of unwitting, different trials and ignore the difference
Several short studies may not anticipate the results of a single study
Comments
Post a Comment