View Poll Results: What is your IS350 1/4 mile time?
13.3
11
36.67%
13.4
4
13.33%
13.5
4
13.33%
13.6
2
6.67%
13.7
3
10.00%
13.8
3
10.00%
13.9
1
3.33%
14.0
2
6.67%
14.1
0
0%
14.2
0
0%
Voters: 30. You may not vote on this poll
Stock IS350 Quarter-mile Data
05-31-11, 07:34 AM
#1
Driver School Candidate
Thread Starter
Stock IS350 Quarter-mile Data
Hey guys,
As part of my AP statistics course final project, I am investigating the statistical significance of Lexus's claimed 1/4 mile time for the IS350 by comparing it to real world data. It would be great if all of you could participate in a poll to supply me with the data for my assignment. Thanks a bunch and vote away!!!
As part of my AP statistics course final project, I am investigating the statistical significance of Lexus's claimed 1/4 mile time for the IS350 by comparing it to real world data. It would be great if all of you could participate in a poll to supply me with the data for my assignment. Thanks a bunch and vote away!!!
05-31-11, 01:23 PM
#6
Driver School Candidate
Thread Starter
Thats right, all I'm interested in is your opinion on 1/4 mile times for 350's.
@ lil4cyl: thanks for the tip
Keep voting! The more feedback I get, the better my assessment will be, thanks!
@ lil4cyl: thanks for the tip
Keep voting! The more feedback I get, the better my assessment will be, thanks!
Trending Topics
05-31-11, 06:40 PM
#8
AP statistics classes usually require you to perform a Student's t test at minimum for a "final project." It's the simplest statistical test available.
There's nothing you can feasibly do with an online voting poll (which does not constitute even a simple "random" sample) and a single 0-60 or 1/4 mile figure provided by Lexus.
Let's assume for a second that this online voting poll did constitute a true random sample. You're asking for 1/4 mile times from stock IS350s, right?
In the end, the only thing you'll get is a (presumably) Gaussian distribution where you can compute things like mean, median, etc. "Statistical significance" has no meaning at all in this context. You're not changing a single variable to see if the effect's of that variable produce a statistically significant (at some designated confidence interval set a priori, usually the 95% confidence interval) change. Now, I don't think you'll even get a Guassian distribution from this online poll, but that's the least of your worries right now.
Here's an example of a paradigm where the concept of "statistical significance" is intelligible -
You measure 1/4 mile times for a stock IS350 (keeping all potentially confounding variables like temperature, humidity, driver, etc. constant) and accumulate the following data -
Stock 1/4 mile times: 13.35, 13.44, 13.26, 13.67, 13.77, 13.82, 13.55, 13.63, 13.64
Next, you install a special (non-stock) intake and you want to see if the intake results in a statistically significant difference in 1/4 mile times.
IS350 with intake 1/4 mile times: 13.22, 13.43, 13.33, 13.10, 13.05, 13.42, 13.16
Now you have two populations and you've constructed your experiment such that only a single variable has changed (stock intake versus, say, F-sport intake). Of course, in practice, it's not possible to fully eliminate all potentially confounding variables, but we do the best we can.
Then you get into issues of one tailed versus two tailed t tests. Since you would expect 1/4 mile times to be lowered, you might use a one tailed t test where rejecting the null hypothesis is easier (by a factor of 2, in fact).
Anyways, what exactly do you have in mind for your project, because what you've said thus far ("investigating the 'statistical significance' of Lexus' claimed 1/4 mile time to real world data") is unintelligible in light what "statistical significance" actually entails.
Maybe I'm wrong and maybe the only point of the project is to test your ability to manufacture random samples, but your method (a clublexus poll) falls flat even under this paradigm.
There's nothing you can feasibly do with an online voting poll (which does not constitute even a simple "random" sample) and a single 0-60 or 1/4 mile figure provided by Lexus.
Let's assume for a second that this online voting poll did constitute a true random sample. You're asking for 1/4 mile times from stock IS350s, right?
In the end, the only thing you'll get is a (presumably) Gaussian distribution where you can compute things like mean, median, etc. "Statistical significance" has no meaning at all in this context. You're not changing a single variable to see if the effect's of that variable produce a statistically significant (at some designated confidence interval set a priori, usually the 95% confidence interval) change. Now, I don't think you'll even get a Guassian distribution from this online poll, but that's the least of your worries right now.
Here's an example of a paradigm where the concept of "statistical significance" is intelligible -
You measure 1/4 mile times for a stock IS350 (keeping all potentially confounding variables like temperature, humidity, driver, etc. constant) and accumulate the following data -
Stock 1/4 mile times: 13.35, 13.44, 13.26, 13.67, 13.77, 13.82, 13.55, 13.63, 13.64
Next, you install a special (non-stock) intake and you want to see if the intake results in a statistically significant difference in 1/4 mile times.
IS350 with intake 1/4 mile times: 13.22, 13.43, 13.33, 13.10, 13.05, 13.42, 13.16
Now you have two populations and you've constructed your experiment such that only a single variable has changed (stock intake versus, say, F-sport intake). Of course, in practice, it's not possible to fully eliminate all potentially confounding variables, but we do the best we can.
Then you get into issues of one tailed versus two tailed t tests. Since you would expect 1/4 mile times to be lowered, you might use a one tailed t test where rejecting the null hypothesis is easier (by a factor of 2, in fact).
Anyways, what exactly do you have in mind for your project, because what you've said thus far ("investigating the 'statistical significance' of Lexus' claimed 1/4 mile time to real world data") is unintelligible in light what "statistical significance" actually entails.
Maybe I'm wrong and maybe the only point of the project is to test your ability to manufacture random samples, but your method (a clublexus poll) falls flat even under this paradigm.
05-31-11, 06:44 PM
#9
And furthermore, you haven't even specified whether you want people to round up or down to the nearest tenth of an integer.
And, I'm not sure it's even appropriate to round data to the nearest tenth of an integer in this case. Imagine if you were trying to derive a random sample of heights and you asked people to round up or down to the nearest foot (up if you are >=.5 down if you are <.5). You'd get data like this: 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 6, 6, 6, 6, 6, 6, 7
Anyways I don't mean to be rude but I don't quite see the makings of a "project" here. You won't have anything to work with. What are your null and alternative hypotheses? Again, "statistical significance" makes absolutely no sense given the paradigm you've provided so far.
And, I'm not sure it's even appropriate to round data to the nearest tenth of an integer in this case. Imagine if you were trying to derive a random sample of heights and you asked people to round up or down to the nearest foot (up if you are >=.5 down if you are <.5). You'd get data like this: 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 6, 6, 6, 6, 6, 6, 7
Anyways I don't mean to be rude but I don't quite see the makings of a "project" here. You won't have anything to work with. What are your null and alternative hypotheses? Again, "statistical significance" makes absolutely no sense given the paradigm you've provided so far.
06-01-11, 08:20 AM
#10
too many factors come in, this is something I did over a year ago:
I did this a couple of weeks ago, I just didnt have the time to post this.
Thanks to Caymandive I was able to get some information form his drag strip journal and use it to answer some questions that i had about drag racing in the IS350... So using a a statistical software i ran some tests.
the main variable taken into consideration was density altittude (DA)..... I wanted to know, not only if DA affects 1/4 mile times(because it obviously does), but by how much. and what times to expect on 1/4 mile on my IS350 according to the weather conditions.
It is hard to explain what density altittude is, but it is basically an international standard of measuring the density of the air, humidity, elevation, and altimeter, all into one number. the lower the density altittude the faster the car will be (basically). usually the lower the temperature is the lower DA is.
So I ran a regression, which is a statistical analysis that lets you forecast 1/4 mile times (in this case) and answer other questions.
From caymandive's journal, I took all the 1/4 mile times and their respective Density Altutudes at the time of the runs. I only took IS350's I did not take IS-F times. otherwise this test would be pointless.
Here are the results:
Regression Analysis: 1/4 mile time versus DA
The regression equation is
1/4 mile time = 13.4 + 0.000137 DA
Predictor Coef SE Coef T P
Constant 13.3584 0.0258 516.78 0.000
DA 0.00013677 0.00002112 6.48 0.000
R-Sq = 53.8%
Analysis of Variance
Source DF SS MS F P
Regression 1 1.0086 1.0086 41.93 0.000
Residual Error 36 0.8659 0.0241
Total 37 1.8745
So, what do these numbers mean:
• Question: Does density altitude have an effect on ¼ mile times?
Yes, Density altitude does has an effect on ¼ mile times. Density altitude has a positive effect on ¼ mile times, which means that as density altitude goes up, so do ¼ mile times. And as density altitude goes down, so do ¼ mile times.
There is only five percent chance that we have answered this question wrong, so we are 95% sure that these statements are true.
• 53.8% of the variation in ¼ mile times can be explained by density altitude. The other 46.2% is explained by other factors such as: track conditions, driver’s skill level, weight reductions, tire psi, quality of tires, gas, head wind, tail wind, etc.
So basically, 53.8% of your 1/4 mile time you can say its due to Density Altittude. the other 46.2% you can say, well this track sucks, or Im not a good driver, or my tires are old. blah blah, etc.
• What times can I expect by looking at a specific DA number?
With density altitude at 0 feet, the Lexus IS350 should run 13.4 1/4 mile E/T time.
For every 1000 feet change in DA, there is a .13 second change in ¼ mile times.
Example: You should run 13.40 @ 0 feet of DA.
You should run 13.27 @ -1000 feet of DA.
You should run 13.53 @ 1000 feet DA.
This is how you can estimate your 1/4 mile time before going to the track by looking at the density altittude.
example: if density altittude is -700......
you use the regression equation = 1/4 mile time = 13.4 + 0.000137 DA
= 1/4 mile time = 13.4 + .000137(-700) = 13.30 E/T time
So, basically with Density Altitude at -700ft. I should run 13.3 on the 1/4 mile. anything lower than that is a good time.
Important NOTE: remember most of these runs were done with intake/exhaust, and spare tire and jack removed. this will give the model some kind of error. But also remember mods dont do much for the IS350.
I did this a couple of weeks ago, I just didnt have the time to post this.
Thanks to Caymandive I was able to get some information form his drag strip journal and use it to answer some questions that i had about drag racing in the IS350... So using a a statistical software i ran some tests.
the main variable taken into consideration was density altittude (DA)..... I wanted to know, not only if DA affects 1/4 mile times(because it obviously does), but by how much. and what times to expect on 1/4 mile on my IS350 according to the weather conditions.
It is hard to explain what density altittude is, but it is basically an international standard of measuring the density of the air, humidity, elevation, and altimeter, all into one number. the lower the density altittude the faster the car will be (basically). usually the lower the temperature is the lower DA is.
So I ran a regression, which is a statistical analysis that lets you forecast 1/4 mile times (in this case) and answer other questions.
From caymandive's journal, I took all the 1/4 mile times and their respective Density Altutudes at the time of the runs. I only took IS350's I did not take IS-F times. otherwise this test would be pointless.
Here are the results:
Regression Analysis: 1/4 mile time versus DA
The regression equation is
1/4 mile time = 13.4 + 0.000137 DA
Predictor Coef SE Coef T P
Constant 13.3584 0.0258 516.78 0.000
DA 0.00013677 0.00002112 6.48 0.000
R-Sq = 53.8%
Analysis of Variance
Source DF SS MS F P
Regression 1 1.0086 1.0086 41.93 0.000
Residual Error 36 0.8659 0.0241
Total 37 1.8745
So, what do these numbers mean:
• Question: Does density altitude have an effect on ¼ mile times?
Yes, Density altitude does has an effect on ¼ mile times. Density altitude has a positive effect on ¼ mile times, which means that as density altitude goes up, so do ¼ mile times. And as density altitude goes down, so do ¼ mile times.
There is only five percent chance that we have answered this question wrong, so we are 95% sure that these statements are true.
• 53.8% of the variation in ¼ mile times can be explained by density altitude. The other 46.2% is explained by other factors such as: track conditions, driver’s skill level, weight reductions, tire psi, quality of tires, gas, head wind, tail wind, etc.
So basically, 53.8% of your 1/4 mile time you can say its due to Density Altittude. the other 46.2% you can say, well this track sucks, or Im not a good driver, or my tires are old. blah blah, etc.
• What times can I expect by looking at a specific DA number?
With density altitude at 0 feet, the Lexus IS350 should run 13.4 1/4 mile E/T time.
For every 1000 feet change in DA, there is a .13 second change in ¼ mile times.
Example: You should run 13.40 @ 0 feet of DA.
You should run 13.27 @ -1000 feet of DA.
You should run 13.53 @ 1000 feet DA.
This is how you can estimate your 1/4 mile time before going to the track by looking at the density altittude.
example: if density altittude is -700......
you use the regression equation = 1/4 mile time = 13.4 + 0.000137 DA
= 1/4 mile time = 13.4 + .000137(-700) = 13.30 E/T time
So, basically with Density Altitude at -700ft. I should run 13.3 on the 1/4 mile. anything lower than that is a good time.
Important NOTE: remember most of these runs were done with intake/exhaust, and spare tire and jack removed. this will give the model some kind of error. But also remember mods dont do much for the IS350.
Last edited by juice14; 06-01-11 at 08:33 AM.
06-01-11, 08:33 AM
#11
Lead Lap
Join Date: Sep 2002
Location: NC
Posts: 407
Likes: 0
Received 0 Likes
on
0 Posts
AP statistics classes usually require you to perform a Student's t test at minimum for a "final project." It's the simplest statistical test available.
There's nothing you can feasibly do with an online voting poll (which does not constitute even a simple "random" sample) and a single 0-60 or 1/4 mile figure provided by Lexus.
Let's assume for a second that this online voting poll did constitute a true random sample. You're asking for 1/4 mile times from stock IS350s, right?
In the end, the only thing you'll get is a (presumably) Gaussian distribution where you can compute things like mean, median, etc. "Statistical significance" has no meaning at all in this context. You're not changing a single variable to see if the effect's of that variable produce a statistically significant (at some designated confidence interval set a priori, usually the 95% confidence interval) change. Now, I don't think you'll even get a Guassian distribution from this online poll, but that's the least of your worries right now.
Here's an example of a paradigm where the concept of "statistical significance" is intelligible -
You measure 1/4 mile times for a stock IS350 (keeping all potentially confounding variables like temperature, humidity, driver, etc. constant) and accumulate the following data -
Stock 1/4 mile times: 13.35, 13.44, 13.26, 13.67, 13.77, 13.82, 13.55, 13.63, 13.64
Next, you install a special (non-stock) intake and you want to see if the intake results in a statistically significant difference in 1/4 mile times.
IS350 with intake 1/4 mile times: 13.22, 13.43, 13.33, 13.10, 13.05, 13.42, 13.16
Now you have two populations and you've constructed your experiment such that only a single variable has changed (stock intake versus, say, F-sport intake). Of course, in practice, it's not possible to fully eliminate all potentially confounding variables, but we do the best we can.
Then you get into issues of one tailed versus two tailed t tests. Since you would expect 1/4 mile times to be lowered, you might use a one tailed t test where rejecting the null hypothesis is easier (by a factor of 2, in fact).
Anyways, what exactly do you have in mind for your project, because what you've said thus far ("investigating the 'statistical significance' of Lexus' claimed 1/4 mile time to real world data") is unintelligible in light what "statistical significance" actually entails.
Maybe I'm wrong and maybe the only point of the project is to test your ability to manufacture random samples, but your method (a clublexus poll) falls flat even under this paradigm.
There's nothing you can feasibly do with an online voting poll (which does not constitute even a simple "random" sample) and a single 0-60 or 1/4 mile figure provided by Lexus.
Let's assume for a second that this online voting poll did constitute a true random sample. You're asking for 1/4 mile times from stock IS350s, right?
In the end, the only thing you'll get is a (presumably) Gaussian distribution where you can compute things like mean, median, etc. "Statistical significance" has no meaning at all in this context. You're not changing a single variable to see if the effect's of that variable produce a statistically significant (at some designated confidence interval set a priori, usually the 95% confidence interval) change. Now, I don't think you'll even get a Guassian distribution from this online poll, but that's the least of your worries right now.
Here's an example of a paradigm where the concept of "statistical significance" is intelligible -
You measure 1/4 mile times for a stock IS350 (keeping all potentially confounding variables like temperature, humidity, driver, etc. constant) and accumulate the following data -
Stock 1/4 mile times: 13.35, 13.44, 13.26, 13.67, 13.77, 13.82, 13.55, 13.63, 13.64
Next, you install a special (non-stock) intake and you want to see if the intake results in a statistically significant difference in 1/4 mile times.
IS350 with intake 1/4 mile times: 13.22, 13.43, 13.33, 13.10, 13.05, 13.42, 13.16
Now you have two populations and you've constructed your experiment such that only a single variable has changed (stock intake versus, say, F-sport intake). Of course, in practice, it's not possible to fully eliminate all potentially confounding variables, but we do the best we can.
Then you get into issues of one tailed versus two tailed t tests. Since you would expect 1/4 mile times to be lowered, you might use a one tailed t test where rejecting the null hypothesis is easier (by a factor of 2, in fact).
Anyways, what exactly do you have in mind for your project, because what you've said thus far ("investigating the 'statistical significance' of Lexus' claimed 1/4 mile time to real world data") is unintelligible in light what "statistical significance" actually entails.
Maybe I'm wrong and maybe the only point of the project is to test your ability to manufacture random samples, but your method (a clublexus poll) falls flat even under this paradigm.
haha... Stupid thin air...
The following users liked this post:
r377 (10-06-21)
06-06-11, 03:59 PM
#13
Driver School Candidate
Thread Starter
too many factors come in, this is something I did over a year ago:
I did this a couple of weeks ago, I just didnt have the time to post this.
Thanks to Caymandive I was able to get some information form his drag strip journal and use it to answer some questions that i had about drag racing in the IS350... So using a a statistical software i ran some tests.
the main variable taken into consideration was density altittude (DA)..... I wanted to know, not only if DA affects 1/4 mile times(because it obviously does), but by how much. and what times to expect on 1/4 mile on my IS350 according to the weather conditions.
It is hard to explain what density altittude is, but it is basically an international standard of measuring the density of the air, humidity, elevation, and altimeter, all into one number. the lower the density altittude the faster the car will be (basically). usually the lower the temperature is the lower DA is.
So I ran a regression, which is a statistical analysis that lets you forecast 1/4 mile times (in this case) and answer other questions.
From caymandive's journal, I took all the 1/4 mile times and their respective Density Altutudes at the time of the runs. I only took IS350's I did not take IS-F times. otherwise this test would be pointless.
Here are the results:
Regression Analysis: 1/4 mile time versus DA
The regression equation is
1/4 mile time = 13.4 + 0.000137 DA
Predictor Coef SE Coef T P
Constant 13.3584 0.0258 516.78 0.000
DA 0.00013677 0.00002112 6.48 0.000
R-Sq = 53.8%
Analysis of Variance
Source DF SS MS F P
Regression 1 1.0086 1.0086 41.93 0.000
Residual Error 36 0.8659 0.0241
Total 37 1.8745
So, what do these numbers mean:
• Question: Does density altitude have an effect on ¼ mile times?
Yes, Density altitude does has an effect on ¼ mile times. Density altitude has a positive effect on ¼ mile times, which means that as density altitude goes up, so do ¼ mile times. And as density altitude goes down, so do ¼ mile times.
There is only five percent chance that we have answered this question wrong, so we are 95% sure that these statements are true.
• 53.8% of the variation in ¼ mile times can be explained by density altitude. The other 46.2% is explained by other factors such as: track conditions, driver’s skill level, weight reductions, tire psi, quality of tires, gas, head wind, tail wind, etc.
So basically, 53.8% of your 1/4 mile time you can say its due to Density Altittude. the other 46.2% you can say, well this track sucks, or Im not a good driver, or my tires are old. blah blah, etc.
• What times can I expect by looking at a specific DA number?
With density altitude at 0 feet, the Lexus IS350 should run 13.4 1/4 mile E/T time.
For every 1000 feet change in DA, there is a .13 second change in ¼ mile times.
Example: You should run 13.40 @ 0 feet of DA.
You should run 13.27 @ -1000 feet of DA.
You should run 13.53 @ 1000 feet DA.
This is how you can estimate your 1/4 mile time before going to the track by looking at the density altittude.
example: if density altittude is -700......
you use the regression equation = 1/4 mile time = 13.4 + 0.000137 DA
= 1/4 mile time = 13.4 + .000137(-700) = 13.30 E/T time
So, basically with Density Altitude at -700ft. I should run 13.3 on the 1/4 mile. anything lower than that is a good time.
Important NOTE: remember most of these runs were done with intake/exhaust, and spare tire and jack removed. this will give the model some kind of error. But also remember mods dont do much for the IS350.
I did this a couple of weeks ago, I just didnt have the time to post this.
Thanks to Caymandive I was able to get some information form his drag strip journal and use it to answer some questions that i had about drag racing in the IS350... So using a a statistical software i ran some tests.
the main variable taken into consideration was density altittude (DA)..... I wanted to know, not only if DA affects 1/4 mile times(because it obviously does), but by how much. and what times to expect on 1/4 mile on my IS350 according to the weather conditions.
It is hard to explain what density altittude is, but it is basically an international standard of measuring the density of the air, humidity, elevation, and altimeter, all into one number. the lower the density altittude the faster the car will be (basically). usually the lower the temperature is the lower DA is.
So I ran a regression, which is a statistical analysis that lets you forecast 1/4 mile times (in this case) and answer other questions.
From caymandive's journal, I took all the 1/4 mile times and their respective Density Altutudes at the time of the runs. I only took IS350's I did not take IS-F times. otherwise this test would be pointless.
Here are the results:
Regression Analysis: 1/4 mile time versus DA
The regression equation is
1/4 mile time = 13.4 + 0.000137 DA
Predictor Coef SE Coef T P
Constant 13.3584 0.0258 516.78 0.000
DA 0.00013677 0.00002112 6.48 0.000
R-Sq = 53.8%
Analysis of Variance
Source DF SS MS F P
Regression 1 1.0086 1.0086 41.93 0.000
Residual Error 36 0.8659 0.0241
Total 37 1.8745
So, what do these numbers mean:
• Question: Does density altitude have an effect on ¼ mile times?
Yes, Density altitude does has an effect on ¼ mile times. Density altitude has a positive effect on ¼ mile times, which means that as density altitude goes up, so do ¼ mile times. And as density altitude goes down, so do ¼ mile times.
There is only five percent chance that we have answered this question wrong, so we are 95% sure that these statements are true.
• 53.8% of the variation in ¼ mile times can be explained by density altitude. The other 46.2% is explained by other factors such as: track conditions, driver’s skill level, weight reductions, tire psi, quality of tires, gas, head wind, tail wind, etc.
So basically, 53.8% of your 1/4 mile time you can say its due to Density Altittude. the other 46.2% you can say, well this track sucks, or Im not a good driver, or my tires are old. blah blah, etc.
• What times can I expect by looking at a specific DA number?
With density altitude at 0 feet, the Lexus IS350 should run 13.4 1/4 mile E/T time.
For every 1000 feet change in DA, there is a .13 second change in ¼ mile times.
Example: You should run 13.40 @ 0 feet of DA.
You should run 13.27 @ -1000 feet of DA.
You should run 13.53 @ 1000 feet DA.
This is how you can estimate your 1/4 mile time before going to the track by looking at the density altittude.
example: if density altittude is -700......
you use the regression equation = 1/4 mile time = 13.4 + 0.000137 DA
= 1/4 mile time = 13.4 + .000137(-700) = 13.30 E/T time
So, basically with Density Altitude at -700ft. I should run 13.3 on the 1/4 mile. anything lower than that is a good time.
Important NOTE: remember most of these runs were done with intake/exhaust, and spare tire and jack removed. this will give the model some kind of error. But also remember mods dont do much for the IS350.
06-06-11, 04:22 PM
#14
Driver School Candidate
Thread Starter
Thanks for you concern and you are absolutely correct syzygy, but the data collected through this poll was good enough to perform the basic level tests required by this project. Don't forget, this is high school and doesn't require real world (realistic) experiments. It plays more of a role as a test to see whether we have mastered the process behind conducting a statistical test, disregarding proper sampling; as this would be way too demanding of a high school student.
Thread
Thread Starter
Forum
Replies
Last Post
