OLDaily

Research Ethics of Twitter for MOOCs
Eamon Costello, Enda Donlon, Mark Brown, Online Learning Journal, 2019/09/03

It should go without saying that I want to do my research ethically. That said, it seems to me that this paper (18 page PDF) on ethics when using data from Twitter goes overboard in its advocacy for caution. After all, what are the privacy implications of a message you have written and posted publicly to several million people? And surely there are as many arguments to be made against anonymizing Twitter posts as are to be made in favour. Collecting tweets isn't like collecting survey data or activity reports, and to treat them as similar seems to me to be wrong. This again I think shows that while so much of 'ethics in research' presumes that a certain ethical code applies, there is not in fact universal agreement on ethics, and even a basic principle such as 'do no harm' can be interpreted any number of ways.

Web: [Direct Link] [This Post]

Is this the right way to use machine learning in education?
Graham Attwell, Pontydysgu, 2019/09/03

Graham Attwell points to ‘Predicting Employment through Machine Learning‘ by Linsey S. Hugo  and comments that it "confirms some of my worries about the use of machine learning in education." The idea is that student records could be used to predict a student's performance on employers' desirable skills. However, say Attwell, "it reduces the purpose of degree level education to employment." Additionally, "it accepts that employers call the shots through proxies based on unquestioned and unchallenged 'well recognised skills'" which may disguise bias against certain social groups. All true, except, how is this different from the current system, other than that it uses technology to do this?

Web: [Direct Link] [This Post]

The Enigma Machine
Jeremy Ashkenas, Observable, 2019/09/03

This is a lovely Notebook example of the Enigma Machine used to encode messages during the Second World War. It's written in Observable, described as the Javascript version of Jupyter Notebooks. What I really like is how it shows how a rotating physical device could be used for encryption. A lot of modern cryptography uses a similar approach, using modulo functions instead of a steel tube (think of it as doing mathematics with a clock). Here's the theory, in cartoon form. In cryptography, each rotor could be thought of as a (base 26 modulo) function, with a salt defining how far each rotor is turned after each input digit.

Web: [Direct Link] [This Post]

Pubfair – A Framework for Sustainable, Distributed, Open Science Publishing Services
Confederation of Open Access Repositories, 2019/09/03

The Confederation of Open Access Repositories (COAR) is calling for fomments on this white paper (13 page PDF). "This white paper outlines a conceptual model for a distributed, international scholarly communication framework, called Pubfair." The idea is to build publishing functionalities over open repositories such as arXiv. In many ways the architecture reminds me of the personal learning environment, though Pubfair also builds in roles for institutions, societies and funders. " Pubfair is designed as an open source software toolkit that can be integrated with existing systems."

Web: [Direct Link] [This Post]