cognitive fun!

Learn your mind. Play it too.
The Reading Span Task (at last!)
cognitivefun | 12 years ago Reply Link me
The Reading Span Task (at last!)
The reading span task (RST) is a common task used to test working memory and reading comprehension. Published by Meredith Daneman and Patricia Carpenter in 1980 (Individual Differences In Working Memory And Reading), it has been widely cited in and adapted for investigations of working memory, reading comprehension, and cognitive processing. The newly added RST follows the first experimental setup within the paper. The original paradigm involves participants reading or listening to a series of two to six sentences and recalling, in the correct order, the last word for each of the sentences read. The number of words one is able to correctly recall is called one's "reading span."

The reading span task is quick to show individual variations in reading efficiency: it was found to correlate strongly with subsequent tests involving reading comprehension: pronoun reference tasks and answering factual questions; it was also moderately correlated with Verbal SAT scores, which involves a fair amount of reading comprehension as well. (A pronoun reference task involves identifying what a pronoun (him, she, that) refers to, typically several sentences ago.)

While the original oral reading span task had participants read sentences aloud, this is very difficult to administer from a browser, so the design here relies on re-typing the sentences instead. Re-typing should be able to guarantee that all words are read before advancing to the next sentence. We all have experience with mindlessly transcribing sentences, but lexical processing most likely still happens during the act -- whether one likes it or not! This is because reading is an automatized process, which means that one has become familiar enough with some action (like reading a word) such that just a tiny nudge (displaying a word) would be enough to trigger the whole waterfall of processes for its completion.

For those into "speed reading" of text: this is why subvocalization (or silently reading to oneself in the head) is so hard to get rid of! Even if your visual field says "shapes!," your phonological loop will still say "sounds!" So assuming automaticity of reading text while typing, this test should give a decent approximation of the reading span, as the RS in oral, silent reading, and listening setups are highly correlated with each other.

As far as strategies go, there isn't a standard "recommended" or "discouraged" strategy. Participants in the original paper used different strategies, but the goal is the same (and difficult!). I will not elaborate on the strategies here just so you can find your own, should you choose to develop one. As for a reasonable speed in typing out the sentences, it turns out that speed does not matter: participants reading at their own pace performed similarly under time-controlled settings. This also goes against the idea that some proponents of reading techniques have, that reading faster allows the brain to more quickly synthesize material, leading to greater comprehension. An effective reader should have good comprehension, regardless of reading speed. Needless to say, effective reading is a learned skill.

Finally, the initial texts for this task are currently extracted from various literary works on Project Gutenberg. If you would like to suggest additions or modifications to the corpus, feel free to write in.
cognitivefun | 12 years ago Reply
Why doesn't this test keep the span I have reached for later trials? As soon as I leave the page, it reverts to 2-span, making it rather tedious to go through all the way to what may or may not be my maximum RS. The option to decrease and increase up to the reached span could be added, like that one found on most tests here (e.g., dual n-back, corsi block test, etc.).

Incidentally, my keyboarding could improve using this rather delightful test. I recommend (and how many more!) as another addition to the text-set. Let us have only entertaining and prized works. ;^)
argumzio | 11 years ago Reply
It's working for me.

Some of the tests only remember the longest span that you have reached if you submit a score for this new level. If not, they offer you the span below on resumption.
cevapcici | 11 years ago Reply
Tho' you probably already know that...
cevapcici | 11 years ago Reply
What you say is true, but not in the case of the Reading Span Task. I have to stay at that page and plug away to the higher levels, and as soon as I leave the page I have to begin from the bottom again.

An additional issue that I noticed but was too tired to mention earlier: the task does not go beyond four items. Once I reach five, the lit-box states "undefined".

Considering that I would like to find my max RS, I would certainly love to have this fixed.
argumzio | 11 years ago Reply
Scratch that. Firefox made a fool out of me again.
argumzio | 11 years ago Reply
My Reading Span is limited by the sheer tedium of bashing out my umpteenth decontextualised sentence or, having typed it out, being unable to find the one typo that prevents it from being accepted.
cevapcici | 11 years ago Reply
Well, now I can share my dunce cap with you. Let us relax in the corner of stupidity.
argumzio | 11 years ago Reply
cevapcici | 11 years ago Reply
Yes, these Blessed Isles and enigmatic smiles.

argumzio | 11 years ago Reply
Doesn't work, cannot submit....
isisdelgado | 11 years ago Reply
Some things are misbehaving and I'm fixing them now.
cognitivefun | 11 years ago Reply
The RST has been fussy about data for a long time but should be alright now. Apparently it affected Mac users more because the Mac uses more precise timing.
I will try to rescue some of the old data as well.
cognitivefun | 11 years ago Reply
Will this ever be fixed?
? | 9 years ago Reply

Login to save scores

© 2008-2012 | about | widgets | blog | cognitive neuroscience for everyone