2015年11月23日 星期一

Revised design of experiment

I checked the list of papers, and I found that they can be divided into three major categories:

  1. Eye tracking (Little combination on Stroop test, designing for special learning disability and other mechanism-enhanced training, binding practice)
  2. Learning style and Dyslexia (mostly logitudinal research, little linguistic or syntax study. Mostly perspectives are from Special education. The research are quite out of date even the latest still rephrase the old finding about defecits on phonology and classroom helping and sympathy to them. )
  3. Synesthesia and (mostly with some statics with little new findings in them; categorization of types)

So I decided to make bridges to un-gap them .
Here are some with useful guidances and findings. From my practical experience in tutoring the dyslexic children, I found it helpful to teach them abstract syntax so that they would understand the semantic meaning of words do not perfectly matching the syntax requirement. 


I can design a mechanical device and teaching procedure  (mnemonic device on syntax structure - POS and some simple rules on it) to trigger the developmental synesthesia on syntax and other senses. I did find it very helpful for them to gain awareness in syntatic structure.
Then we put these children to the eye tracking machines to see the pattern of their eye movement. (Stroop test, before and after the training. That shall be )

As for learning styles, I am pretty sure it did help when tutoring the children. The older brother is obviously more analytic thinking, while the younger sister is a visual learner. She draws comic girl characters all the time. When I tutored them, I gave them different tasks to complete.
The girl needs more attention and care since she's more psychological dependent (as the youngest of the family)
The boy is

For three months and hours of teaching, they have gained quite some advances in legitimate sentences making.
Their predecessor teachers never make advances like these for over ten years. I had tested the boy's ability in sentence making on learned structures. He's good, showing a real comprehension on rules on each grammar compounds. I was worrying if it was me boosting myself on the effect of teaching, and it turns out a NO.

But I need the eye tracking techs to see the solid evidence on lessen regression and fixation. The POS and mnemonic device might be the reason of improvement on their understanding structures. For the perspectives of "neighborhood size", we know that POSs have certain rules to be serially built.

For example, just give a fake numbers for elaboration:
N initiate POSs have the following types (if you know the HMM, you can make calculation on how fast and efficient it would be to grouping POSs to larger units. )
N + V is 75%  in the copora
N+  adv +V is 23%
N+ adv + adv + V is 2%

Let's do the math:
You can predict that the possibility of succedding POS [V] right after N is far higher generally. You can speed up the reading by checking only the important part of the sentences-that is, the predicates (and the core of them). This would be a chain of multiplication and finally we will know that certain POS combination dominates the whole corpora. (according to the writing style)

We know the structure could be well-embedded, but once the dyslexics students obtain the tricks of segmenting and analyzing, they could do it well. Even for non-analytical learners like the girl, they can still benefit from the tricks. 

Yet, I need evidence. I am not sure if I will find them in the parafoveal processing or regression (looking for missing POS groups). So far, I can do little in designing these. I can only assume we will find difference before and after the training. (Yes, we may need several weeks to do the post-test. So that they would not going back to the old ways of parsing.)
As for the perspective of synesthesia, I think it would be way too hard to initiate it. There might be some other element or critical period reasons. But I did successfully demostrated the linear order of a SVO language - English. The legitimate structure can be correspond to the different layers of plates. Showing hidden ideas on "tense", "perfections",  
 
Sorry I am in great headache, I got fever and my blood pressure soars like rocket. So this is the brief version of experiment design.

Eye Movement Reading experiment design Draft

My professors:




This might be an intruding letter, I made some integration on: the dial plate, synesthesia, dyslexia, Eye tracking, learning style, HMM analysis of POS and so on.
It has a wide span on different genres of research. If you are not interested in this experiment, I would not send it again.
For those from the perspectives of syntax: see E1
For those from the perspectives of eye tracking, see E1, E2, E3, E4
For those from the perspectives of HMM model, see E2, E3
For those from the perspectives of mechanical and dial design, see E1
For those from the perspectives of TESOL and vocabulary teaching, see E1, E4
I got two main reference books in hand, and a bunch of eye tracking and dyslexia / synesthesia papers.
Most importantly, I got two real dyslexics children as my tutoring students. I am a synesthete actually. I am helping them with the dial and teaching method I design. 
The designing of the the experiment leads to a question: can man-triggered synaesthesia make dyslexics read and comprehend better?
I was inspired by and influenced by the following papers.  
  1. Bergfeld-Mills, C.,Howell Boteler, E.. & Oliver, G.K. (1999). Digit synaesthesia: A case study using a stroop-type test. Cognitive Neuropsychology, 16(2), 181-191.
  2. Baron-Cohen, S., Harrison, J., Goldstein, L.H., & Wyke, M. (1993). Coloured speech perception: Is synesthesia what happens when modularity breaks down? Perception, 22, 419-426.
  3. Besner, D. (2001). The myth of ballisitc processing: Evidence from Stroop's paradigm. Psychonomic Bullentin and Review, 8, 324-330.
  4. Besner, D., & Stolz., J.A. (1999b). What kind of attention modulates the Stroop effect? Psychonomic Bulletin and Review, 6, 99-104. 
  5. Chinn, S.J. & Ashcroft, R. (2006) Mathematics for Dyslexics: Including Dyscaculia, 3rd edn. London: Wiley.
  6. Gray, J.A. (1999). The hard question on consciousness: Information processing versus hard wiring. In C. Taddeo-Ferretto & C. Musio (Eds.), Neurinal bases and psychoogical aspects of consciousness, vol. 8. (pp. 450-457). Singapore: World Scientific. 
  7. Grossenbacher, P.G., & Lovelace, C.T. (2001). Mechanisms of synesthesia: Cognitive and physiological constraints. Trends in Cognitive Science, 5, 36-41.
  8. Hochstein. S., Ahissar, M. (2002). View from the top: Hierarchies and reverse hierarchies in the visual system. Neuron, 36(5), 791-804.
  9. Humphreys, G.W., Cinel, C., Wolfe. J., Olson, A., & Klempen, N. (2000). Fractionating the binding process: Neuropsychological evidence distinguishing binding of form from binding of surface features. Vision Research, 40, 1569-1596.
  10. Lakoff, G. (1987). Women, fire, and dangerous things: What categories reveal about the mind. Chicago: University of Chicago Press.
  11. Luria, A. (1968). The mind of mnemonist. New York: Basic Books.
  12. MacLeod, C.M. (1991). Half a century of research on the Stroop effect: An integrative review. Psychological Bulletin, 109(2), 163-203.
  13. Mattingley, J.B., Rich A.N., & Payne, J. (2003). The effects of attentional load on colour-graphemic synesthesia. Manuscript submitted for publication. 
  14. Mattingley, J.B., Rich, A.N., Yelland. G., & Bradshaw, J.L. (2001b). Unconscious priming eliminates automatic binding of colour and alphanumeric form in synesthesia. Nature, 410, 580-582.
  15. Pashler, H. & Johnson, J.C.(1998). Attentional limitations in dual-task performance. In H. Pashler (Ed.), Attention (pp. 155-189). East Sussex: Psychology Press.
  16. Shadlen, M.N., & Movshon, J.A. (1999). Synchrony unbound: A critical evaluation of the binding hypothesis. Neuron, 24, 67-77.
  17. Singer. W., & Gray. C.M. (1995). VIsual feature integration and the temporal correlation hypothesis. Annual Review of Neuroscience, 18, 555-586.
  18. Smilek, D., Dixon, M.J., Cudahy, C., & Merikle, P.M. (2002). Synesthetic color experiences influence memory. Psychological Science, 13, 548-552.
  19. Treisman, A. (1996). The binding problem. Current Opinion in Neurobiology, 6, 171-178.
  20. Treisman, A. (1999). Solutions to the binding problem: Progress through controversy and convergence. Neuron, 24, 105-110.

To answer the question, I set up a series of  experiment to see if all these hypothesis would work.
  • E1: Pre-test. Picking up suitable reading materials. Before and after training, read syntatic similar articles.  [record pattern of regression and saccade. ]  根據我的陽春轉盤設計,大約是最基礎的5 layers,也就是S+V+det/adj+O 的句型。用這個訓練一陣子,或是最傳統的,請學童把主詞和動詞圈出來就好。經過一段時間的練習,他就能發覺主語和述語的開頭端點在哪裡。轉盤是為了把抽象的POS具現化分離出來,以及用材質和振動去刺激synesthesia的產生。(有點像是遊戲手把的振動回饋)
  • E2: Mid-test. After training, "lessen neighborhood size of POS." [record pattern of regression and saccade. ] 
  • E3: Post-test, replacing the lexicons in the previous articles. Keep the syntatic structure. [check for pattern of regression and saccade. ]
  • E4: checking result and if the prediction is right (there shall be some eye movement pattern change after the training. I don't know what it might be. But it would be definitely some changes.) [compare the data in pre, mid, and post test.]
"attention allocation"
Prof. Tsai 提到的,閱讀速度的加快可能是來自其他訊息,比 如對context的背景知識了解,或是隱藏的temporal coding 訊息。(I think it's the small neighborhood size of POS combination, indicating the syntatic structure of sentences.  這比較能解釋為什麼對於不同領域的技術文件或是不熟悉的詞彙,還是能保有比一般人還要快的processing。skipping for unfamiliar words. )

基本上我對閱讀速度和效率的提升,定義為更快的 parsing和理解(quickly get the causality among the participants in the sentences. )
所以我設計S和O的部分做區隔,V的部分也做區隔。
"neighborhood size" is a helpful definition for my theory on geometrical grammar. I can finally see the reason why it enhance the overall performance of reading to me.
In short, there're only limited possibilities to see the combination of POS. 
在"parafoveal processing in reading"一文中,提到「the efficiency of reading can be greatly increased if the reader start to identify the upcoming parafoveal word before they fixate it」

這樣就不再是用上學期思考的HMM(Hidden Markov Model 隱性馬可夫鍊)的方式去思考POS 或句構組成的機率問題。畢竟,人腦和電腦不一樣,對於機制的運作,一定有一個更好更真實的描述狀況。眼動的 neighborhood size搭配上HMM可能會是一個很符合真實運作方式的解釋模型。
I was looking for explanation from the perspectives of synesthesia, See the pictures, The letter H (or A) is decided by higher level of cognitive abilitiy, so the people with synesthesia would report mix feeling on the color of the letter. I believe the same thing happens on lexical parsing on POS. 
We all know syntax-semantic continuity. The parsing begin when there's clue for semantic meaning incoming, then we do the syntactic analysis. There might be some regression, but I think the fixation in post-test (stucked, processing on hard words) would be far less compared to pre-test. 

=======
There're some other interesting topics worthy of reading. But for now I will just work on related papers and contact the researchers. 
To put it simple, I choose to monitor how students look for the syntatic verb and subject in an essay. The material is specifically designed for similar experiment from last year's Gestalt experiment. The same thing modified by the articles I wrote. 
Check for visual search (they are thinking syntax is separated from semantic, which is wrong. This is a common mistake and phenomenon for beginners)

A long-term experiment design would be like this: put the dial on the screen, and give the user a controller (like something for PS4 or xbox 360)  to control what they see on the screen. I might need to make an app to make the objects move according to the input of controller.