Originally Posted by DuckieTigger
That doesn't surprise me at all, given the results and basing them on 10 individuals. That is probably 10 that are used to reading electronic devices all day long. It doesn't mean anything other than: If you look long enough and pick the right people, you can "prove" anything.
And design the study to tweak things...
I am not
saying this study was designed that way. But if I personally wanted to demonstrate no significant difference when a real difference exists, as well as keeping the sample size low, I'd do things like set up older e-ink devices under sub-optimal lighting in an imperfect font, and compare that to brand new retina-screen tablets with the most readable fonts and the brightness, backgrounds and contrast tweaked just perfect. And then write the methodology section carefully to conceal most of this.