This is the code I wrote in R to merge, clean and analyse the data from 80 participants that took part in my first Python-based computer experiment.
Click here to see the code for data cleaning.
Click here to see the code for data analysis.
I am doing a PhD in the analysis of human behaviour, largely through the use of computer experiments. Without bogging you down with archaic technical details, I generally study how people 'make sense of things' and particularly how they respond when things don't make sense.
To do this I use computer experiments built with the open-source Python-based software PsychoPy (Peirce, 2007; http://www.psychopy.org/about/overview.html). Due to the technical and novel nature of my experiment this did require quite a bit of coding with Python, though coding isn't always necessary on that platform. These experiments look like modern brain-training tasks such as the SMART (Strengthening Mental Abilities with Relational Training - https://raiseyouriq.com/what_is_smart).
In the first experiment for my PhD I designed a training procedure that would easily allow me to manipulate 'relational networks' to ensure that some 'make more sense' than others. From this, we can easily compare how people respond to networks that make sense, compared to networks that don't.
The variables I measured were numerous, including: reaction time, response choice, response consistency, affect, arousal, and self-reported sense-making.
A PhD is a 3-7 year long research project in which a dedicated, organised, perseverant individual aims to build a research programme that is written up as a thesis. Far removed from undergraduate study, the PhD candidate is the project manager, taking full responsibility for the long-term research project, organising the research, managing and organising the supervisors of the project, and many more activities that would easily be found in most research-active companies and organisations. The PhD candidate comes up with a research project, devours all the literature in that area over a period of usually 6 months, and takes copious organised notes to summarise the field. This forms the literature review. Following this, an experiment is devised, designed, piloted, goes through ethical application procedures, and gets run with hundreds of participants over hundreds of research hours. Next, the data will be cleaned and prepared for analysis, which is what I do in the first linked script above. Then the data will be analysed, turned into boiled down, technical insights and beautiful graphs (see the second linked code script above), and written up as a thesis chapter. The next experiment then builds on the findings of the previous chapter, and so on. I expect to have around 7-9 PhD studies complete by thesis hand-in. Once the thesis is handed in, external examiners read the ~100,000 word document and you have a viva, where essentially the examiners aim to poke holes in your work whilst you defend it, much the same way that you might defend work pitched in a firey board meeting.
All in all, the PhD requires someone able to study something in minute scientific detail over a period of several years. Candidates need perseverance, optimism, organisation, and a bit of ingenuity. Ultimately, it demonstrates the ability to work in a technical, research-oriented and statistical environment, and is the highest educational achievement available in the UK system.
Released under the GPLv3+ open source license.