Ratio: The Simple Codes Behind the Craft of Everyday Cooking.
Michael Ruhlman has a new book and it’s about the place where math and food meet. In other words, he wrote it especially for me.
Comments
I want this.
This could be the catalyst for me finally getting into cooking.
Ha, speaking of math, I talked to our newer neighbors for the first time yesterday. I suspect two of them are in high school. I kinda want to ask if either of them is in a statistics class, because I never took it myself. I want to read the textbook.
I’ve really wanted to understand how to determine and apply standard deviation lately.
Sammy said:I’ve really wanted to understand how to determine and apply standard deviation lately.
MS Excel: =STDEV
My entire graduate program has been based on statistical analysis. It’s pretty cool, and you’d be amazed at how powerful Excel is.
Okay well, for now, simply put, standard deviation tells us how far away most of the data are away from the mean/average. And by most, I mean 68% of all stats.
Paris said:Okay well, for now, simply put, standard deviation tells us how far away most of the data are away from the mean/average. And by most, I mean 68% of all stats.
Yeah, I guess the Excel formula doesn’t really explain what exactly is going on, eh?
Jay said:Paris said:Okay well, for now, simply put, standard deviation tells us how far away most of the data are away from the mean/average. And by most, I mean 68% of all stats.
Yeah, I guess the Excel formula doesn’t really explain what exactly is going on, eh?
Don’t worry: Neither does what Paris said.
That’s not what I mean at all. I could’ve looked up the definition, but didn’t. I was more interested in how to calculate standard deviation on my own, which is what Jay said to a degree… but I was looking to do it a little more manually.
The thing is, we have a program at work that tests our kids to see what “levels” they’re on. That’s all fine and good. It’s pretty straight forward, and there’s a chart you can reference to figure it all out. But when I went to the seminar teaching the administration (I was invited on a whim) how to use this knowledge, the spokesman said, as an example, that if your class has an overall standard deviation above a certain number, like 10, whole group instruction was impossible. As in, you can’t just teach to the class and expect them to get it.
I think our Standard Deviation was like 16.
Well okay then, calculating is something like finding the average of how far away each entry is from the actual average.
- Make a list of each value (and find the mean)
- Find the difference between each value and the mean.
- Square each difference (this makes everything positive).
- Average those differences.
- Calculate the square root of that number, and there you have it!
Paris said:Well okay then, calculating is something like finding the average of how far away each entry is from the actual average.
- MS Excel: =STDEV
Thanks!
Math!
Sammy said:But when I went to the seminar teaching the administration (I was invited on a whim) how to use this knowledge, the spokesman said, as an example, that if your class has an overall standard deviation above a certain number, like 10, whole group instruction was impossible. As in, you can’t just teach to the class and expect them to get it.I think our Standard Deviation was like 16.
Well, unless you guys are talking about some sort of standardized scale (which you probably are), the standard deviation on it’s own doesn’t mean much. You need to have some sort of notion about the order of magnitude of the mean for the STDEV to tell you anything. I’m assuming you are working on a 0-100 grading scale, in which case a STDEV of 10 is going to be pretty significant (10 is big relative to any mean that might fall in that range).
The way you can apply it to your job is as follows. I assume your goal is to teach to the mean (average , not the Jesus Nieves looking little fucker) student. If the mean student is getting an 80, statistics show that 68% of your students will then score between 64 and 96 (80 +/- 16). In other words, too many of them are failing if you teach to the mean, so you’re fucked.
On the other hand, let’s say you’re working with a sensor of some sort that measured super high temperatures (mean of 1000 degrees or something). A standard deviation of 10 degrees is small in relation to the mean, so that’s actually pretty good.
Just in case some of you are curious as to the psychometrics we’re talking about, here’s the normative data from which we’re drawing conclusions (in pdf form!).
So it goes beyond a 0-100 type scale. It’s 0-something else, with “benchmarks” along the way type of deal. There’s another sheet somewhere on the internet that gives a probability of passing the state test dependent on their scores on this. For example, if my kid gets a 203 in Math, there’s a probability of 43% that they’ll score as “Proficient” on the state test.
Most of my kids are around 203. The government continues to fail to realize that Special Ed students working below level should not be expected to be proficient at grade level. That’s like asking a guy with one leg to enter an ass kicking contest, sans prosthetic.
Hooray for applying business models of standardized performance assessment to wholly individualized and highly varied personal and intellectual growth! Hooray for compulsory schooling!