Graeme Blake, creator of LSAT Hacks, taking questions Forum

[akg144]

Hey Graeme first of all thanks alot for answering our questions I really appreciate you doing this.

Now my question pertains to Reading Comprehension. I'm really struggling in that section and it's the only thing holding me back from a great LSAT score. For the past 5 PTs I've got 96.7 correct on LG, 85% on LR but only 77.5 right on Reading Comprehension. If I could get -3 or -4 on RC instead of 6.2 wrong per RC section it would make a huge difference. So my question is this:

There appears to be two major approaches to LR:

1) Read the stimulus in about 60-80 seconds for only 2 purposes 1) to get the main point question right and 2) to star the authors point of view throughout the passage in preparation for those 2-3 questions about authors view. The rest of the questions which require specific details can then be answered in the remaining 6 minutes and change by referring back to the text.

2) Take 3-4 minutes and mark the passage extensively and then "breeze through the questions" (only problem is I never breeze through) and allocating only 30/question to answer allowing for minimal reference back to the text (3-4 minutes for the 6-8 questions has always been a difficulty for me so I've usually leaned towards the former rather than the latter).

My question is this: I've tried both with minimal success. How do you approach reading comprehension? Did you ever struggle with it? If so what did you do to improve? I usually get the main point and authors attitude very well but no matter what method I try I usually just end up getting the other questions wrong usually 3 or 4 per science passage about 2 or 3 wrong on the other passages without much variation. I've tried just about everything circling key phrases such as however or on the other hand, boxing in main characters, starring the author, tagging the paragraphs, reading predictively, doing the Manhattan Scale everything but I just can't seem to improve on RC. Any advise would be greatly appreciated!

Thanks Graeme!

[LSAT Hacks (Graeme)]

Graeme,

I saw on a couple of your previous threads where you said that improving a significant number of points from one's diagnostic is possible but very rare (talking like 150 to 170). Obviously this is a shared view on TLS. I just have kind of a general question. Do you think this is because achieving this high a jump takes a tremendous amount of work and most people just don't do it, or another reason? Thanks for your help.

That view is correct. The problem is that that jump takes more than work. If hard work were the only requirement, then everyone would get 170+

To go from 150 to 170 takes hard work, intense analysis, and possibly even getting smarter.

Around 160-165 I see a dividing line. Below that line, people routinely mix up formal logic. Above that, they almost never do, except for a rare misread.

I'm not suggesting that formal logic is enough to get a high score (it's not actually that important), but rather, that some kind of mental shift tends to happen above and below ~162, and formal logic is a sign of that shift.

160-170 is a similarly big gap. Almost no one gets a 170. That's a very, very tough score to get.

I see students improve all the time from ~140 to 155, no problem. The lower down you are, the easier the points are to get. Higher up, they're harder.

So, you haven't got an easy road ahead. And remember, this forum is subject to sample bias. You'll see people say "I went from 145 to 170", but you don't hear from all the people who score 147, got up to 153, and gave up after months of hard work.

It can be done, but you'll need to do a lot of work to figure out why you're making mistakes, and how to stop making them. Just doing more questions won't be enough. There are specific things you're doing right now that make you make mistakes. Errors in your thought process. You need to analyze your own mind and root out those errors.

Edit: Of course, the lower your LG score right now, the better. It's the most masterable section. I see no reason anyone reasonably intelligent can't get to -0, with time and effort.

[Louis1127]

Thanks so much, Graeme. I appreciate you taking the time to answer my question so thoroughly. Take care.

[LSAT Hacks (Graeme)]

Hi Graeme -

First off, thanks for hosting this thread! I am currently preparing to retake in October, and am PT'ing in the high 160s - 170s. I would ideally like to PT around 175-180 before October. I have taken a handful of practice tests since June, using exams that I have seen before, and am starting to build a list of questions that I got wrong. Given that the exams I took are not even new to me, the fact that I am still getting questions wrong clearly means I did not review thoroughly enough the first time I used these tests. Do you have any suggestions as to the most effective way to review wrong answers for LR/RC/LG?

I took 7sage before the June exam, so I am familiar with Blind Review and redoing LG 10 times. I have been doing this for each PT I take. However, I still miss anywhere from -2 to -4 on exams that I have seen before. The fact that my average number of incorrect questions per exam has remained fairly stable suggests to me that my reviews are not allowing me to see the underlying patterns necessary to avoid traps from test to test. Therefore, I am wondering if you have any further advice, guidance, or insight into this?

Thanks!

If it makes you feel better, I sometimes get questions wrong on tests that I've *taught* before. Usually, sloppiness is the cause, but a lot depends on your state of mind.

You asked a good question, and I don't have a guaranteed answer. I think Blind Review is a good idea, as is redoing LG.

As you say, that's not enough. Have you tried a study group?

I suspect that on some questions, you're identifying wrong reasons for why an answer is right or wrong. It's very easy to do. Having a study partner forces you to verbalize your ideas. Making thoughts more explicit removes a bit source of error.

Also, a study partner will usually have different errors than you do. So they'll catch where you're going wrong. I find that if 2-3 students who score 150-160 talk about a question, there's very little they don't figure out on their own.

Conversely, it's almost impossible to know if your reasons are right without outside feedback.

Ok, some section specific tips:

LR
-----------

For each argument:

1. Identify the conclusion
2. Identify the reasoning
3. Identify what's wrong with it.
4. Make a post here on TLS in case of doubt about any of these.

For flawed reasoning, review all the wrong answers, and think of examples. Make them concrete. These wrong answers repeat, and they're very hard to understand at first. You should train yourself to understand them.

Sufficient Assumption: Almost all can be solved by diagramming, by connecting the gap between premises and conclusion. On review, make sure you know how to do this.

Parallel Reasoning: Argument structure is crucial. Practice making the correct structure. See how you could have eliminated the wrong answers faster for structural reasons.

LG
---------

Not much to add to repeating + watching explanation videos. There is a hidden, easy path through each game. Find that path.

If a question is hard, that means you're still missing something. Figure out how you could have solved it in 15 seconds.

Practice memorizing the rules. It's not that hard, especially if you draw a couple directly on the main diagram. Almost all LG errors come from forgetting rules.

RC
--------

Almost all non-main point/primary-purpose questions can be answered using specific lines from the passage. Practice finding a specific line to prove or disprove each answer.

I find if students practice this, they can find most lines within 2-3 seconds. You obviously need to have a decent grasp of the passage structure, but you should be doing that anyway.

If you can find line references quickly, then you can go from 90% sure to 100% certain on most questions.

All Sections
----------

Why did you pick the wrong answer? Are there any labels you can give to the mistake?

Saying that you get strengthen questions wrong isn't very useful. There are 1,001 reasons you could get a strengthen question wrong.

Saying that you mixed up percentage and amount is more useful. That error repeats across question types. If you get it wrong once, you will probably make the mistake again unless you work at fixing it.

Hope that helps, just some scattered thoughts.

[LSAT Hacks (Graeme)]

Thanks so much, Graeme. I appreciate you taking the time to answer my question so thoroughly. Take care.

No problem. Don't let me discourage you too much. As long as your scores keep increasing, you're on the right track.

[LSAT Hacks (Graeme)]

By the way, explanations for LSAT 68 are on my site now. I hope to put more up soon, early feedback has been encouraging.

[objection_your_honor]

This question isn't really about LSAT prep, but: how often do you personally take full timed PTs? For all the dissection you do in your books, do you also try to maintain a connection to the complete test-day experience?

[LSAT Hacks (Graeme)]

This question isn't really about LSAT prep, but: how often do you personally take full timed PTs? For all the dissection you do in your books, do you also try to maintain a connection to the complete test-day experience?

I think that's important. I can't take many timed tests because I've either done them all or done parts with students. But each time a new test is out I take it under timed conditions before writing explanations.

I'm considering taking an official LSAT again at some point, to remember what the test center experience is like.

[mvonh001]

By the way, explanations for LSAT 68 are on my site now. I hope to put more up soon, early feedback has been encouraging.

ya they are fantastic thanks

[LSAT Hacks (Graeme)]

Hey Graeme first of all thanks alot for answering our questions I really appreciate you doing this.

Now my question pertains to Reading Comprehension. I'm really struggling in that section and it's the only thing holding me back from a great LSAT score. For the past 5 PTs I've got 96.7 correct on LG, 85% on LR but only 77.5 right on Reading Comprehension. If I could get -3 or -4 on RC instead of 6.2 wrong per RC section it would make a huge difference. So my question is this:

There appears to be two major approaches to LR:

1) Read the stimulus in about 60-80 seconds for only 2 purposes 1) to get the main point question right and 2) to star the authors point of view throughout the passage in preparation for those 2-3 questions about authors view. The rest of the questions which require specific details can then be answered in the remaining 6 minutes and change by referring back to the text.

2) Take 3-4 minutes and mark the passage extensively and then "breeze through the questions" (only problem is I never breeze through) and allocating only 30/question to answer allowing for minimal reference back to the text (3-4 minutes for the 6-8 questions has always been a difficulty for me so I've usually leaned towards the former rather than the latter).

My question is this: I've tried both with minimal success. How do you approach reading comprehension? Did you ever struggle with it? If so what did you do to improve? I usually get the main point and authors attitude very well but no matter what method I try I usually just end up getting the other questions wrong usually 3 or 4 per science passage about 2 or 3 wrong on the other passages without much variation. I've tried just about everything circling key phrases such as however or on the other hand, boxing in main characters, starring the author, tagging the paragraphs, reading predictively, doing the Manhattan Scale everything but I just can't seem to improve on RC. Any advise would be greatly appreciated!

Thanks Graeme!

I combine both approaches. I read in about 90 seconds, understand everything, and refer back extensively.

How the heck can I do that? I read really, really fast. About 600 words per minute.

You can measure yourself with this test: http://www.readingsoft.com

The college-educated average is 200-300 words. We treat reading speed as a given, but it's easy enough to change. I find that students in my classes who actually try the stuff I say can jump 50 WPM in a week or two.

Here's what to do:

1. Measure your reading speed
2. Go to this site: http://www.spreeder.com, click 'click here to spreed this passage' to get started

site settings:

i. Set chunk size to 3 (and look in the middle of the words when reading)
ii. set width to 800px
iii. Set background color to white
iv. Change the speed to 50 above your normal speed
v. Hit save
vi. Press play

3. The site will explain what to do. Afterwards, practice reading articles this way by going to a newspaper, hitting 'print' so you get the full text without pictures, and pasting it into the box that comes up after you hit 'click here to spreed'.

4. Do this every day for 5-10 minutes. Increase it slightly each day, so that you feel a little uncomfortable. Occasionally mix it up by setting the speed far faster than you're comfortable with and see what you can get.

This will force you to reduce subvocalizing. Your brain can understand words faster than you can physically read them, at present.

You should see a boost in your reading speed. Either of the strategies you mentioned amount to saying 'half-ass it'. I think everyone has some room for easy reading speed improvements, which makes RC much more manageable.

Note that I am talking about increasing your baseline comfortable reading speed. I'm not talking about reading faster and understanding less. The goal is reading faster while understanding just as much or more. People often find their comprehension improves with speed.

p.s. The instructions will make more sense when you actually go to the site. And, this method only works if you actually try it. A lot of people nod, listen, try it once and then ignore it. Do it solidly for 2-3 weeks before judging it, and test your reading speed for improvements.

[flash21]

Hey, not sure if you'll know what I'm talking about, but JY ping said it usually takes about 5 PT's for people to hit their potential (he said that on the 7sage site when I was looking at the courses he offered, and some had more PT's than others, the minimum being 5).

Wondering if you could elaborate, if you know what I am even referencing to. Thanks.

[flash21]

Hey G.

I am getting absolutely crushed by SA questions. I rage quit my study session today after getting owned so badly by them, so I am laying in my bed mad as hell right now. Any advice would be appreciated. I plan on just trying to memorize the LRB chapter and drilling.

[LSAT Hacks (Graeme)]

Hey, not sure if you'll know what I'm talking about, but JY ping said it usually takes about 5 PT's for people to hit their potential (he said that on the 7sage site when I was looking at the courses he offered, and some had more PT's than others, the minimum being 5).

Wondering if you could elaborate, if you know what I am even referencing to. Thanks.

I'm actually not sure what you're referring to. You'd have to ask JY. I personally recommend many more PTs than that, and I think he does too.

I am getting absolutely crushed by SA questions. I rage quit my study session today after getting owned so badly by them, so I am laying in my bed mad as hell right now. Any advice would be appreciated. I plan on just trying to memorize the LRB chapter and drilling.

Sufficient assumption questions are actually one of the most learnable questions types, so don't worry.

They almost all have one of the following structures:

Conclusion: A --> D

Evidence:

1. A --> B --> C
2. A --> B C --> D
3. B --> C --> D

In all cases, you need to link the evidence to the conclusion. So, draw your conclusion on the page, spread apart, like this:

A D

Fill in the evidence (I'm using the first one):

A --> B --> C D

You can see you need an arrow between C --> D, or the contrapositive D --> C

That's it. Almost all sufficient assumption questions follow that simple structure. You just need to practice diagramming, identifying the conclusion and evidence, and aligning them.

Repeating SA questions is very useful, because you're learning a skill, like LG diagramming.

Edit: The post didn't include the spaces I drew. Just imagine a big space wherever there isn't an arrow. I could make a blog post with proper formatting and link to it if that would help.

Does anyone know how to make posts include more than one space between letters?

[flash21]

Thanks for the advice G. Will be doing all your suggestions!

Another question - I currently am scoring a low 150's on my PT's (ive taken just one since my diagnostic which was 140). I am writing in October, do you think I'm on pace to hit 160 (you think I have enough time?)

My breakdown for my last was 15 RC , 15 LR, 13 LR, and 14 LG's (correct)

Thanks.

[flash21]

Also, how much do you charge for tutoring? Sorry for spamming your thread lol.

[objection_your_honor]

44.4.20

P1: EMP winners of the past 25 years are covered by Acme
C: Since EMP has recognized that it offers financial security, it is probably a good plan for anyone with similar needs.

The biggest flaw I'm seeing here is that being covered by Acme does not mean endorsement. The stimulus says the EMP winners have "thus clearly recognized," but they could all be wholly dissatisfied with the plan despite being covered.

I think both (A) and (D) address this flaw.

(A) I'm not comfortable with "majority" and "previous years" here — it seems more general than the scope of the argument. Also, just because the EMP winners have used a retirement plan other than Acme does not mean they endorse that one, or that Acme has lost their endorsement or "clear recognition."

(D) The correct answer, but I'm having a lot of trouble with it. The argument does not assume that the EMP winners have "deliberately selected" the Acme plan. All that's given is that the winners are covered, and from that coverage the argument claims endorsement. That move is the flaw — how coverage came about is of no consequence.

So what's the deal?

[LSAT Hacks (Graeme)]

44.4.20

P1: EMP winners of the past 25 years are covered by Acme
C: Since EMP has recognized that it offers financial security, it is probably a good plan for anyone with similar needs.

The biggest flaw I'm seeing here is that being covered by Acme does not mean endorsement. The stimulus says the EMP winners have "thus clearly recognized," but they could all be wholly dissatisfied with the plan despite being covered.

I think both (A) and (D) address this flaw.

(A) I'm not comfortable with "majority" and "previous years" here — it seems more general than the scope of the argument. Also, just because the EMP winners have used a retirement plan other than Acme does not mean they endorse that one, or that Acme has lost their endorsement or "clear recognition."

(D) The correct answer, but I'm having a lot of trouble with it. The argument does not assume that the EMP winners have "deliberately selected" the Acme plan. All that's given is that the winners are covered, and from that coverage the argument claims endorsement. That move is the flaw — how coverage came about is of no consequence.

So what's the deal?

How coverage came about has EVERYTHING to do with endorsement. The argument is incorrect to say that having the plan = endorsement.

If you choose a plan, you implicitly endorse it as being a plan that suits your needs.

If your employer chooses your plan, then your employer endorses it. You didn't endorse the plan. You don't have a say in which plan your employer chooses.

As for A, it's not persuasive. If EVERY winner from the past 25 years endorsed the plan (assume they did), then that's quite significant. 25 years is a long time, and these are the most recent winners. Who cares if some prior winners used a different plan? That plan might not even exist anymore.

[LSAT Hacks (Graeme)]

Thanks for the advice G. Will be doing all your suggestions!

Another question - I currently am scoring a low 150's on my PT's (ive taken just one since my diagnostic which was 140). I am writing in October, do you think I'm on pace to hit 160 (you think I have enough time?)

My breakdown for my last was 15 RC , 15 LR, 13 LR, and 14 LG's (correct)

Thanks.

Sounds like you're doing alright. Hitting 160 from low 150s is definitely achievable by oct, especially since you still have room to improve with LG. Of course, to be guaranteed a 160, you'll want to get your range up to 161-164, which is harder than simply hitting 160 a few times on timed PTs.

Also, how much do you charge for tutoring? Sorry for spamming your thread lol.

I sent you a PM about tutoring

[itsanemily]

Thanks for all the advice, Graeme!

One other thing I wanted to get your opinion on: do you think it's better to drill a specific question type, or to take preptests, and then review carefully? I take about 1-2 preptests during the week, and then go over then very carefully before starting another one, making sure that I understand why my answers are incorrect. However, I have never tried drilling, although I have heard it mentioned on this forum as a strategy. My mistakes don't seem concentrated in any particular question type, which is why I dont think it would be useful for me. Would be interested to read your thoughts on this, though.

[objection_your_honor]

44.4.20

P1: EMP winners of the past 25 years are covered by Acme
C: Since EMP has recognized that it offers financial security, it is probably a good plan for anyone with similar needs.

The biggest flaw I'm seeing here is that being covered by Acme does not mean endorsement. The stimulus says the EMP winners have "thus clearly recognized," but they could all be wholly dissatisfied with the plan despite being covered.

I think both (A) and (D) address this flaw.

(A) I'm not comfortable with "majority" and "previous years" here — it seems more general than the scope of the argument. Also, just because the EMP winners have used a retirement plan other than Acme does not mean they endorse that one, or that Acme has lost their endorsement or "clear recognition."

(D) The correct answer, but I'm having a lot of trouble with it. The argument does not assume that the EMP winners have "deliberately selected" the Acme plan. All that's given is that the winners are covered, and from that coverage the argument claims endorsement. That move is the flaw — how coverage came about is of no consequence.

So what's the deal?

How coverage came about has EVERYTHING to do with endorsement. The argument is incorrect to say that having the plan = endorsement.

If you choose a plan, you implicitly endorse it as being a plan that suits your needs.

If your employer chooses your plan, then your employer endorses it. You didn't endorse the plan. You don't have a say in which plan your employer chooses.

As for A, it's not persuasive. If EVERY winner from the past 25 years endorsed the plan (assume they did), then that's quite significant. 25 years is a long time, and these are the most recent winners. Who cares if some prior winners used a different plan? That plan might not even exist anymore.

Choosing it may imply endorsement, but that's only one way of many to arrive at endorsement (or "clear recognition," as the argument says). The flaw in the argument is making that leap generally, not necessarily in assuming the winners have "deliberately selected" the plan. You could have your retirement plan assigned to you and yet still "clearly recognize" that it offers financial security. As a result, the particular assumption in (D) is simply not present in the argument.

[LSAT Hacks (Graeme)]

Thanks for all the advice, Graeme!

One other thing I wanted to get your opinion on: do you think it's better to drill a specific question type, or to take preptests, and then review carefully? I take about 1-2 preptests during the week, and then go over then very carefully before starting another one, making sure that I understand why my answers are incorrect. However, I have never tried drilling, although I have heard it mentioned on this forum as a strategy. My mistakes don't seem concentrated in any particular question type, which is why I dont think it would be useful for me. Would be interested to read your thoughts on this, though.

If your mistakes aren't concentrated, then I don't really see the advantage of drilling. I think timed tests are most important.

That said, a few question types have optimal strategies. Parallel reasoning and sufficient assumption are the biggest examples. They often use formal logic, and they're slow if you do them incorrectly. Can be worth drilling and reviewing those in particular.

You should likewise review the wrong answers on flawed reasoning questions extra carefully. They repeat, and they're hard to understand. Get used to reading them, and you'll go faster.

[LSAT Hacks (Graeme)]

44.4.20

P1: EMP winners of the past 25 years are covered by Acme
C: Since EMP has recognized that it offers financial security, it is probably a good plan for anyone with similar needs.

The biggest flaw I'm seeing here is that being covered by Acme does not mean endorsement. The stimulus says the EMP winners have "thus clearly recognized," but they could all be wholly dissatisfied with the plan despite being covered.

I think both (A) and (D) address this flaw.

(A) I'm not comfortable with "majority" and "previous years" here — it seems more general than the scope of the argument. Also, just because the EMP winners have used a retirement plan other than Acme does not mean they endorse that one, or that Acme has lost their endorsement or "clear recognition."

(D) The correct answer, but I'm having a lot of trouble with it. The argument does not assume that the EMP winners have "deliberately selected" the Acme plan. All that's given is that the winners are covered, and from that coverage the argument claims endorsement. That move is the flaw — how coverage came about is of no consequence.

So what's the deal?

How coverage came about has EVERYTHING to do with endorsement. The argument is incorrect to say that having the plan = endorsement.

If you choose a plan, you implicitly endorse it as being a plan that suits your needs.

If your employer chooses your plan, then your employer endorses it. You didn't endorse the plan. You don't have a say in which plan your employer chooses.

As for A, it's not persuasive. If EVERY winner from the past 25 years endorsed the plan (assume they did), then that's quite significant. 25 years is a long time, and these are the most recent winners. Who cares if some prior winners used a different plan? That plan might not even exist anymore.

Choosing it may imply endorsement, but that's only one way of many to arrive at endorsement (or "clear recognition," as the argument says). The flaw in the argument is making that leap generally, not necessarily in assuming the winners have "deliberately selected" the plan. You could have your retirement plan assigned to you and yet still "clearly recognize" that it offers financial security. As a result, the particular assumption in (D) is simply not present in the argument.

Ah, I see what you're saying. Let's break down the argument's evidence:

1. The winners are covered
2. They thus endorse it

If the winners chose the plan, then 2 follows naturally. Why choose a plan you don't endorse?*
If the winners didn't choose the plan, they MIGHT still endorse it. But it's no longer a certainty based on the evidence.

If none of the winners chose the plan, then the automatic link between evidence and conclusion is broken.

To prove an argument wrong, you don't have to prove the conclusion false. You just have to show the evidence doesn't lead to the conclusion. You said you "could" have the plan assigned and yet recognize it offers financial security. "Could" is not good enough.

* There are a few warranted assumptions here. We can assume that a winner of the most prestigious economic prize has at least some reasonable level of income. Prestigious economists aren't poor. We can also assume that prestigious economists choose retirement plans that will work well for them.

Therefore, if a prestigious economist chooses a plan, it's a tacit endorsement.

[kiyoku]

Graeme, I thank you as well for the time you put into this.

I have a question about the speed reading that you suggested. I've tried the couple of URLs you posted and I remember trying this in the past for a week or so. I assume I didn't try for long enough to truly benefit from this.

I understand many people talk about eliminating vocalization and sub-vocalization when reading to increase speed since saying a word takes longer than reading it without saying it.
Somehow this seems extremely hard for me. What I find interesting is that you suggested that I increase the speed about 50wpm higher than what my normal reading speed is.

My normal speed (when trying to be on the faster end) is only 200 wpm. This is a huge problem for me.
Now here's my question:

Should I be doing the exercise in an attempt to not sub-vocalize at all? Is the objective to get rid of the sub-vocalizing habits and see how much I can retain while reading in a completely different form?
(because if this were the case, i feel i'll spend 2 weeks blanking out in front of the screen looking at words go by... in other words, 0% sub-vocalization for me is nearly 0% comprehension).
(also if this is the case, then I find it necessary to bump the wpm to something at least above 350wpm or else I start sub-vocalizing some of the words).

Or did you suggest that I increase it by 50wpm higher so that I ease into the process of firstly REDUCING subvocalization and then moving on to slowly ELIMINATING it.

Or did i just miss the whole picture?

[LSAT Hacks (Graeme)]

Graeme, I thank you as well for the time you put into this.

I have a question about the speed reading that you suggested. I've tried the couple of URLs you posted and I remember trying this in the past for a week or so. I assume I didn't try for long enough to truly benefit from this.

I understand many people talk about eliminating vocalization and sub-vocalization when reading to increase speed since saying a word takes longer than reading it without saying it.
Somehow this seems extremely hard for me. What I find interesting is that you suggested that I increase the speed about 50wpm higher than what my normal reading speed is.

My normal speed (when trying to be on the faster end) is only 200 wpm. This is a huge problem for me.
Now here's my question:

Should I be doing the exercise in an attempt to not sub-vocalize at all? Is the objective to get rid of the sub-vocalizing habits and see how much I can retain while reading in a completely different form?
(because if this were the case, i feel i'll spend 2 weeks blanking out in front of the screen looking at words go by... in other words, 0% sub-vocalization for me is nearly 0% comprehension).
(also if this is the case, then I find it necessary to bump the wpm to something at least above 350wpm or else I start sub-vocalizing some of the words).

Or did you suggest that I increase it by 50wpm higher so that I ease into the process of firstly REDUCING subvocalization and then moving on to slowly ELIMINATING it.

Or did i just miss the whole picture?

The 50 WPM suggestion is because you can naturally read faster on spreeder than on paper, when you set it to three words at a time.

Just yesterday I tested a student's normal reading speed, it was 300 WPM. I had them read the spreeder instructions at 350 WPM, it was easy for them.

I always tell students "If the instructions seem too fast or too slow, pause and adjust them". So if you try 250 and are bamboozled, just set it lower. +50 WPM is just tendency, not a hard law. The goal is just to find a starting pace you're comfortable with.

You shouldn't be actively trying not to subvocalize. That's what the software is for. Once you're comfortable with a speed, increase it the next day. As it goes faster, you'll naturally subvocalize less. It's the only way to keep up.

The reason this works is because your brain is capable of understanding words faster than 200 WPM. You're just not physically or mentally capable of sending words to your brain faster at the moment. This software will get the words there faster, and you'll be able to understand.

You can also throw in a few sessions of words much too fast, like 400 or 600 WPM. The goal is not to understand everything, just pick up what you can.

I recently had a student jump from 230 to 300 in a week using these methods. I may make a video guide for this. It's one thing I find incredibly easy to explain in person or on Skype, but I find my text explanations never seem as clear.

[kiyoku]

Graeme, Thanks so much. I'm far more interested in the idea that I try to push my speed slowly instead of putting the focus on reducing subvocalization. I've tried speed reading techniques before and I never really felt that they were useful for any sort of rigorous reading so I was having a difficult time understanding exactly what you meant.

I just have one more question regarding this reading exercise you suggested.

I assume that you can just copy and paste any text that you want in there. What kind of text do you usually prefer to put in for your students?