My question is concern with the correct categorization of the data collection phase
As known in the research methodology phases, there is a data collection phase .
I want to write the process of the data collection phase :
Is it right to put the process :
Data Selection and Data Preparation Or
Data Preparation and Data Preprocessing and put the selection of the data as part of the data preparation ?
Looking Forward For your answers
Thanks,
Dear Sumaia
The data collection phased usually consists of multiple steps (all of which you might wish to report on in any publication):
- Definition of your population
- Decision on the sampling strategy
Once you have the data you would perform:
- Data cleaning
- Data wrangling
- Data analysis
I assume that with data selection, you likely mean data cleaning and with data preprocessing, data wrangling?
I hope this is in some way helpful to you. This is certainly a very simplified perspective on it, but still, could serve as a starting point for you.
All the best for your research.
Best,
Daniel
I would go with the the latter. Data preparation and processing and let data selection be part of that process.
It is a combination. As Daniel mentioned: it's a process of multiple steps.
In some situations you collect data from existing systems (for example from transactional systems like administration) to analyse (asking questions afterwards, see what's in there).
In other situations you can start the data collection process after asking specific questions. The data collection is a step in the research project itself. For example using Experimental Design, which let's you generate a specific data matrix before starting the data collection itself (asking questions before to create a specific data set).
Most data is not ready for analysis from scratch and need data preprocessing (especially when coming for sources that were not intended for analysis or coming from multiple sources where data were collected in different formats). In the latter case you need to recode and adjust data.
In most research books a simple data set is analysed. In more advanced situations you may run different analyses on different parts of a combined data set (or use subsets). Sometimes even during analysis you may need to alter the structure of the data set (for example flipping cases and variables).
Another nice example of data preprocessing and preparation during analysis and data collection is with text mining when you wish to recode a text document to numeric variables. First you have to know what words are important (analysis), then the text mining procedure creates a numeric summery, that can be used in additional analyses in combination with data from other sources.
In data mining you quite often have a flow of multiple analyses where the output from one procedure is used as input for the next one (but may be preprocessed).
So the order of the steps may differ between situations. Both data preprocessing and preparation are inherent parts of the research project.
Thanks a lot for your answers Wim Nieuwenhuijsen Sachin Suknunan Daniel Dauber Dilraj Kaur
- Badges
- Science topic
- Similar topics
- Methodology
- Research Methodology