I am in concurrence with the remarks of Dr.Kamal, which is very valid and honest, you have to be open and straightforward to avoid any unpleasant occurrence.
Yes, You can use your own data and cite your first publication in data description section. In fact, you can any public data or your own data many times.
I have seen one case that the author made his data public and his article in cited over 100 times .
The new article should be different in methodology and results section and should present something new and different from your previous one. It seems that EFA (Exploratory Factor Analysis) and CFA (Confirmatory Factor Analysis) are similar approaches.
Yes, Rahul. Doing so is called SECONDARY analysis or RE-analysis. You use the data to answer a different research question.
I agree with all of the above. Yes, you can use your data and especially if you have many independent variables not tested in the first study.
Yes, sure. You can use your own data for different aims provided you refer to the data of the first study.
Feasibility study should be carried out with the "required data" with that of the "Available data"... if positive towards the aim of establishing a culture of data sharing and archiving.
This contribution should be closely looked into for issues of analysis.
You can use the data by giving the reference to the first. obviously second article should conclude something new.
Yes, of course. Research methodology accepts new conclusions from second-analysed data base. You must just refer to the original research correctly.
Dear All,
Old data can be reused if one can attain thorough them a new consequence or another set of reasoning and results. Reuse of old data can show the development of a researcher.
When different old data sets are recombined in a new analysis (e.g. data from local studies lumped together to study wider-scale patterns) you can confirm or not confirm conclusions from local studies.
In many long-term studies the data from particular years are 'used' and 'reused' and 'reused' etc... Interesting is that how it comes that the same data can be placed in so many different theoretical frameworks as is often the case for long-term studies.
Many thanks to you all. It was your great support.
Your answers are not great support only but apt to the context in which the discussion was held.
Thanks a lot again!
I agree that you can reuse the data for different purposes provided you refer to the data published before. However, it is not recommended to run EFA and CFA with the same sample, it is always better to use two independent samples.
Generally, using the same data sets for both EFA and CFA is not considered good. Reason is that factor structure derived from EFA is likely to fit the data well when the same data is used in CFA analysis. Better approach is to split the data into two, then using EFA on one half and CFA on the other half.
I agree with Ridhi, that the data set should be divided into two separate sets and the two techniques should be used on different sets.
As we know that the exploratory factor analysis is a totally data dependent technique, it involves a number of subjective decisions and should be succeeded by confirmatory factor analysis (CFA) to cross validate the structure. If we see, the basic purpose of CFA is to check whether the factor structure matches with the results of the original study. In this case, if the same data set is used to conduct CFA, then we are bound to overlook the testing part, instead we are intentionally providing such data to the software which will fit the structure.
If in case it is not feasible to collect data again or in case the research does not have a large data set then another approach could be-
Check for the number of measured variable (items in the questionnaire) taken and divide the data in two parts by maintaining 5:1 ratio for one part and keep rest in the other. 5:1 ratio means 5 records (or responses) for every variable taken i.e. if there are 300 responses and total 25 variable are taken then keep 25*5= 125 responses in one part and rest 175 in other. Take 125 responses for EFA study and rest 175 for CFA study. Though the more is the number for CFA study, the better it is but still you will be able to maintain the statistical requirement of the technique.
- Badges
- Science topic