Monthly Archives: October 2013

Japanese Guest Houses

This site is a well constructed way of finding ryoken or other traditional Japanese accommodation. A lot of Japanese travel websites are only available in Japanese or are hard to navigate and find your way around – this site is a refreshing change. It also has a fantastic insight guide – my favourite part of which is the guide to unforgivable things to avoid doing in Japan – wearing shoes inside and getting soap in a bath are the main two.

Link: Japanese Guest Houses

What’s going on when I push my belly out?

Ask Science is a joy, it’s a place you can go and ask scientists anything and get well written answers with sources. It’s part of reddit which is generally a cesspit but Ask Science is carefully curated with throwaway comments, jokes and unsourced answers deleted. I like the way this simple answer quickly develops into a more complicated question about how breathing works.

Link: What’s going on when I push my belly out?

First Rule of Usability? Don’t Listen To Users

Jakob Nielsen has been providing great advise for interaction design for years. This article is useful if you’re considering doing a survey on your website. It isn’t as black and white as the title suggests – you can ask users things that relate to their immediate action (ask ‘Why did you visit this site?’ just after they arrive) but in general you should try and look at what users actually do, not what they say they want.

Link: First Rule of Usability? Don’t Listen To Users

Warning Signs in Experimental Design and Interpretation

This is an article I keep coming back to by Peter Norvig, currently Google’s Head of Research. On a weekly basis I see news reports on scientific papers with the journalist wildly misinterpreting the study (usually with the aid of the exaggerated press release from the lab that published the paper). It helps to find the original paper if it is available, or read the press release, with this article as a guide to warning signs. As Peter says it doesn’t tell you an the results are incorrect but it gives you a warning to look further.

Link: Warning Signs in Experimental Design and Interpretation