New researchers usually face the issue of finding the gap in the methodology which they can exploit in the construction of new models. This blog intends to provide few insights regarding the ways in which the change in existing literature of methodology can be introduced. Following are few dimensions which can be explored.
- Change at variable level
New studies can use different ways to introduce novelty in the model at the variable level.
1.1. It can be in terms of different proxy of the variable
1.2. It can be a new index of all available proxies
1.3. It can be subcomponent of a variable (i.e. cyclic portion, volatility portion)
1.4. It can be in term of its units
1.5. It can be different form of the variable (i.e. log form, reciprocal form)
- Change at the specification level
New studies can use different specification of the variables to create novelty
2.1. Quadratic version of IVs can be used
2.2. Cross product of different IVs can be used
2.3. Splitting of the variable into increasing and decreasing portion
2.4. Study of ordinal differences in the dependent variable
- Change at the data level
Several forms of data can be used to modify or expand the scope of research by increasing sample size
3.1. Cross-sectional data
3.2. Time series data
3.3. Panel data (static panel or dynamic panel)
- Modification in the estimation approach
In this blog, several estimation models have been discussed which can be used. Here the use of assumptions can play a crucial role in the determination of the model.
- Engineering a model
Lastly, a model can be radically improved if the change is incorporated at all levels of the model. In this stage, two or three different models can be engineered to make a new model.