Methodology

AllStation Laboratory produces applied quantitative research at the intersection of environmental science and mathematics, with some policy sprinkled in between. Each project takes a clear question about an Indian environmental situation (including energy, infrastructure, urban systems, etc.) and builds a mathematical analysis to answer it.

This methodology is similar to exploratory modelling and analysis. The goal to understand how a system behaves, and to quantify outcomes that are often discussed in vague, unspecific terms. The scope is as such: ASL studies questions that are quantifiable, India-specific, and policy-relevant. ASL may or may not explore policy implications in detail, but providing a direction to policy it certainly is a goal of the papers.

ASL projects are not peer-reviewed. However ASL does aim for full transparency: all data sources are cited, all assumptions and limitations are stated explicitly, and all methodology is available in the paper appendices. An outline of the python code used for analysis can also be found in the appendices.

Every ASL analysis follows the following consistent structure.

ASL relies on secondary data from peer-reviewed academic literature, government datasets, and established news and research organisations. Some projects may dive into information provided by private corporations (especially for infrastructure topics, where data may not be peer-reviewed). All sources are appropriately cited in APA format.

For example, typical data sources include publications from the Ministry of New and Renewable Energy (MNRE), the Central Electricity Authority, peer-reviewed journals, and established policy research organisations. Whenever assumptions or extrapolations are involved, they are stated explicitly.

ASL models are intentionally parsimonious, a concept the author learnt while learning mathematical modelling as a course — they are kept as simple as the question allows, possibly sacrificing some complexity in favour of clarity and interpretability. This is a choice which keeps the model and interpretations straightforward. All assumptions and limitations caused by the same are stated in the paper.

Every ASL paper states its assumptions explicitly. All inputs are available for independent verification.

ASL does not claim that its models are complete representations of reality. As George Box observed: all models are wrong, but some are useful. The goal is to always clear a vague situation.

All ASL analyses are publicly available. Every paper includes appendices containing a derivation of mathematics (when required), data input given to the code (with citations), and full outline of the code itself, allowing the methodology to be independently verified.

The goal of this openness is transparency, alongside demonstrating that rigorous independent research is possible, replicable, verifiable. Scrutiny is invited, and feedback is always welcome.

ASL is committed to being transparent with regards to artificial intelligence, i.e. where it is and is not involved in its work.