Reproducible research is fundamentally important and is not done well in social sciences e.g., here or here or here.
My data and research code is freely available here: https://github.com/pmpk20/
The code for my thesis is pretty horrible, but I promise the more recent scripts are better.
Sometimes estimation objects are missing because they can be really large filesizes.
If you are a learning/working in choice modelling/survey methods and have a questions on why/how I did something, please feel free to email me.
I am gathering as many Stated Preference surveys in one place as I can here: github.com/pmpk20/Surveys
If you can contribute with a survey or questionnaire you've worked on, please send it over
I will put my "intro to surveys" course material here too.
"Willingness-to-pay for precautionary control of microplastics, a comparison of hybrid choice models" (Journal of Environmental Economics and Policy, 2023, https://doi.org/10.1080/21606544.2022.2146757) is available here: https://github.com/pmpk20/PhDHybridChoiceModelPaper
Not particularly nice code be warned.
"Stated preferences for the colours, smells, and sounds of biodiversity" (Ecological Economics, 2025, https://doi.org/10.1016/j.ecolecon.2024.108410) is available here: https://github.com/pmpk20/WinterPaper and here: https://data.kent.ac.uk/480/
Okay this I'm mostly happy with. Some chatgpt-inspired tidying helped.
Scripts numbered according to order of execution.
If some output data is missing that's bc the filesize is insane. I might be able to email it you if need.
"How confident are you in the ability of experts to provide reliable information? Evidence from a choice experiment on microplastics" (Environment and Development Economics, 2025) is publicly available here: https://github.com/pmpk20/PhD_CEPaper
This one starts with cleaned data already, the actual raw survey responses -> cleaned data is in my PhD repo, but the rest of the paper can be replicated from here.
I report session information for the scripts to run on my machine, but let me know if there is some issue.
"Economic valuation of pest regulation benefits provided by arthropods in the UK" (Ecosystem Services, 2025) is publicly available here: https://github.com/pmpk20/DruidD1 and https://doi.org/10.5518/1674
So this is code to simulate an economic production function across different crops and parameters. It's setup to replicate the paper as published but it shouldn't be too difficult to tweak the number of iterations, the yield response, the pest x natural enemy interactions, the input data etc., and other parameters if you want. Let me know if you want to work on it.
Future work should apply this spatially.
"Heterogeneous Preferences for Use and Non-use Values Provided by Insects." (People and Nature, 2026) is publicly available here: https://github.com/pmpk20/DruidD2 and https://doi.org/10.5281/zenodo.20799844
Final Ngene design, survey design, R code, choice modelling data + output
We recovered respondent-level post areas (e.g., LS1) but this is removed from the repo.
Added a Claude-derived codebook of each variable from the survey and analysis.
Be very careful of hitting 'run' on scripts 07+08 - I used High Performance Computing to execute them!
"Quantifying temporal changes in the monetary value of forest cultural ecosystem services globally: a meta-analysis". (Environmental Research Letters, 2026) is publicly available here: https://github.com/pmpk20/Forest_CES_MetaAnalysis and here: https://doi.org/10.22024/UniKent/01.01.501
Claude Code (Sonnet 5) was amazing at tracing + fixing issues across the whole replication package.
The package includes (i) Data/codebook.csv to help describe the variables we end up with (ii) Data/Data_StepZero.csv with the raw data extracted during the meta-analysis (iii) all the R scripts necessary for data transformation, modelling, and output to match the paper.