Reading on Paper is Faster than iBooks on the iPad

[HarryT]

Folks,

Keep it polite, please. It's fine to disagree with someone's opinion; it is not acceptable to make personal attacks.

Thank you.

[Moderator]

[tompe]

Originally Posted by nekokami

There's a big difference (more than an order of magnitude) between 24 and 350. And I prefer the 5% error margin, so one would need at least 500 participants to design a study, even using these guidelines.

The other problem with these guidelines is that they don't take the complexity of the sample into account.

Yes, of course. But depending on what you are studying and the result a smaller sample size can give statistical significant result. I read the comments about sample size in this thread as a disbelief that a small sample size ever could be statistical significant. And since that is a common error to believe that I just wanted to point that out.

[LDBoblo]

Originally Posted by nekokami

Keep it civil, please.



There's a big difference (more than an order of magnitude) between 24 and 350. And I prefer the 5% error margin, so one would need at least 500 participants to design a study, even using these guidelines.

The other problem with these guidelines is that they don't take the complexity of the sample into account. If you need to control for several variables, your sample size needs to be higher to be sure you've adequately represented everyone. In this case, we should be looking at the type of material being read, the reading ability of the participants, vision issues, language of the materials and primary language of the reader, level of comfort with technology of the reader, reading conditions (e.g. outdoors in daylight vs. in a well-lit room vs. in a dimly-lit room), etc. Essentially, any circumstances you want to be able to generalize across need to be taken into account in your sample selection. (If you only want to be able to generalize to American full-time college students at four-year public institutions, your sample probably wouldn't need to be as large.)

In this case, the results of the comparison were not found to be significant. That's partly because of the small sample size. A study of this size doesn't have sufficient power to detect a significant difference, even if there is one. The qualitative data collected (the participants enjoyed the experience of using an ebook reader) is fine, though not generalizable. A study of this type can be useful to generate hypotheses which could be tested in a larger, more quantitatively rigorous study. But no conclusions can be drawn from the study as presented.

(And for those who wonder about my qualifications, I'm working on a dissertation involving quantitative research in the use of technology in the learning process-- not that far from the stated goals of this study.)

I think what irks me is that the practical limitations of their findings are elucidated in the study, and even in the synopsis, had anyone bothered to read the bloody thing. Instead, we get quite a few rather extreme reactionary posts, most of which are wholly ignorant of even the most basic aspects of the study and its contents.

One would have expected more reading from a forum centered around reading.

[nekokami]

Originally Posted by LDBoblo

I think what irks me is that the practical limitations of their findings are elucidated in the study, and even in the synopsis, had anyone bothered to read the bloody thing. Instead, we get quite a few rather extreme reactionary posts, most of which are wholly ignorant of even the most basic aspects of the study and its contents.

One would have expected more reading from a forum centered around reading.

I can appreciate your frustration, but an ad-hominem attack seems uncalled for, and unlikely to change anyone's mind. Perhaps a well-chosen quote from the synopsis regarding the study's limitations would be more effective here?

[nekokami]

Originally Posted by tompe

Yes, of course. But depending on what you are studying and the result a smaller sample size can give statistical significant result. I read the comments about sample size in this thread as a disbelief that a small sample size ever could be statistical significant. And since that is a common error to believe that I just wanted to point that out.

Yes, but the page you linked to suggested 10% of the total population you wish to generalize to, with at least 30 participants. The sample size doesn't hold up as a quantitative study. It's quite good for a qualitative study, but then it would be inappropriate to be reporting numbers and significance tests.

[astra]

Originally Posted by happy_terd

It would take me longer to finish a book on an iPad. Too many distractions.

One of the main reasons why I don't want all-in-one and am a strong adherent of dedicated ebook readers.

[WT Sharpe]

Originally Posted by LDBoblo

Frankly, I hope they weeded out any people so hopelessly incompetent that they would be unable to turn a page after being taught how to do so.

But there's an art to learning to anticipate how many words beforehand to hit that "next Page" button so the reading experience continues without breaks in the continuity that only comes through experience. As we become more familiar with the speed at which page turning occurs after the button is pressed, compensating for any delay becomes second nature. Someone experienced with their e-reader will naturally out-perform someone who has merely been trained in its use for the sake of an experiment, all other things being equal.

[jbjb]

Originally Posted by tompe

Yes, of course. But depending on what you are studying and the result a smaller sample size can give statistical significant result. I read the comments about sample size in this thread as a disbelief that a small sample size ever could be statistical significant. And since that is a common error to believe that I just wanted to point that out.

That's exactly right - the number of samples required for a given level of statistical significance depends very much on the nature of what is being studied, the nature of its statistical distribution etc.

For example, if we *know*, a priori, that a die has a heavy bias towards a particular number, such that a roll of that die is known to have a 99% chance of being X, where X is an unkown in the range 1 to 6, then we don't have to make many rolls at all to have a very highly statistically significant idea of X.

(Please note that I'm not saying anything about the study in question, just following up on the more general statistical question.)

/JB

[Shaggy]

Originally Posted by dpayment

Studies like this prove nothing really, I call them "cesspool agitation studies", they're usually designed to cause controversy.

..and generate site traffic. That's probably the only reason it was done.

[BenLee]

Originally Posted by WT Sharpe

But there's an art to learning to anticipate how many words beforehand to hit that "next Page" button so the reading experience continues without breaks in the continuity that only comes through experience. As we become more familiar with the speed at which page turning occurs after the button is pressed, compensating for any delay becomes second nature. Someone experienced with their e-reader will naturally out-perform someone who has merely been trained in its use for the sake of an experiment, all other things being equal.

That's actually a great point, which I had not considered previously. I guess because page turning has become so natural when reading paper books, that I hardly even consider that there is actually some processing involved to time the page turn at just the right moment. It's become something that happens subconsciously in the background for most readers. But of course, this needs to also happen for eReaders as well. Although I suspect there is less an art in this case because it is completely about timing, less about maneuvering fingers over the page edges in just the right way and preparing for the turn, etc.