So I have to do a statistics project and I decided to do it on powerlifting, my (if you know statistics) null hypothesis is “Does height affect how much a person can deadlift?” I need to find information about this so if you guys could point me to where I could find such statistics that would be great. Thanks in advance.
n=1
6’5" and 601
I do NOT buy into the theory that taller guys are better built for deadlift, as it has been my slowest progressing lift for nearly my entire training career. Maybe if you have monkey arms and a short torso, possibly, but if you are proportionate and tall, I think it is actually a hindrance.
Thanks for the info but I need at least 40 samples (lifters) so if you know of a website or something that would be great, currently I’m looking up lifters then looking up their heights in wikipedia.
This gon’ be good.
You could pretty much get n=400 by posting an open excel document where people can anonymously post their height and best deadlift. If you’re interested, you can even add as many variables as you like to work with. Example, gear use, chemical assistance, training age, biological age. If you work the results using a statistics software like the SPSS it’s not really that much extra work.
I’m just throwing out some thoughts here. Manually looking up these statistics sounds like a lot of work and probably would bias the results considerably anyway.
Try to find a free survey site where you build your questionaire, then send the link to people, then it will collect and store your data. surveymonkey.com I think is one.
Post the links to a bunch of different forums.
Honestly you could try digging around this site. Just start going down the list of training logs. If they’re not a pussy, you should be able to find a deadlift somewhere in there. A lot of people who don’t have training logs will still have a max deadlift posted in a forum somewhere, so you could go searching through post histories to try and find one. Height is often listed in their profile. I bet you could get 40. You can use me too lol. 5’9 - 525. So using me and VT so far I guess deadlift increases with height lol.
You could create a public Google Doc spreadsheet and allow for public editing. Then post the link here in the PL forum so that anybody reading the thread can enter their data. For example:
I realize this is just a school project, so just do a good job, and who really cares if you can get something meaningful out of the results. That being said, if I had to do something like this for serious, I wouldn’t be able to use “height”. I believe you would have multiple confounding variables (or they might have called them lurking variables or hidden variables). I’m not sure if you’ve studied that yet or not. That’s when you have a variable that is correlated with both your independent (height) and dependent (deadlift) variables. For example, arm length, leg length (upper and lower?), torso length, and hand size come to mind. Those would probably all have multicollinearity issues though.
It would be cool to do actually; I’m sort of geeking out now. I wish I could find some data for it. Then we could see which factors play the biggest role and by what margins and in what combinations do they work best. Could also see how those variables generally relate to height. I would want Ed Coan to be part of my study too. He would be an outlier I bet lol.
64 424
66 385
66 400
67 490
67 551
68 315
68 450
68 495
69 330
69 375
69 500
69 525
70 150
70 210
70 353
70 410
70 480
70 480
70 490
70 500
70 543
70 555
71 462
71 475
71 584
71 625
72 243
72 410
72 487
72 545
73 440
74 418
74 600
74 625
74 815
75 474
75 510
77 380
77 601
78 455
I did this out of curiosity. Here’s 40 people’s heights(in.) and deadlift maxes(lbs.) from TNation’s training logs. (sorry they’re so scrunched; couldn’t fix that on this website)
The regression is absolute shit though. You can add in more data if you like, but based on what I just posted your conclusion is going to be that height and deadlift strength are totally unrelated.
There’s no way you’re going to be able to establish correlation using just height and I’m not sure that lifters here on TN are going to be your best sample in any case. You’ll want actual competition results to know the lift was good and to know that the lifter in question actually trains for that lift. But you’re also going to need to know the age of the lifter, the lifters weight, number of years lifting, etc. Limb length would also be good to know in this case.
What exactly are you trying to do here? Is this for a stats course? I think that using a single factor like this is going to lead you to an incorrect conclusion. Tell me more about this project. If anything all you’re really going to learn is that using the wrong data leads to poor results. I’m not sure you’re really going to learn anything about stats going this route.
james
Just because you might have a negative correlation, or the correlation doesn’t even exist, doesn’t mean it’s a poor choice to be studied. It’s still a result. As well, confounding variables are just fine if you’re looking for a correlation. A correlation =/= cause and effect.
You can learn plenty about stats if there isn’t a correlation. Perfect examples aren’t real life, anyway.
As well, not sure if you posed your null in that way because this is T-Nation and not your course, but your null hypothesis isn’t a question. Your null should be “Increased height has no effect on amount of weight deadlifted”
H0(naught): Height has no effect on amount of weight deadlifted
Your alternate can then be “Increased height has an effect…” if you’re using a two-tailed test, or “Increased height will cause an increase (or decrease, if that’s what you wanna test) …” if you’re using a one-tail test.
[quote]atypical1 wrote:
What exactly are you trying to do here? Is this for a stats course? I think that using a single factor like this is going to lead you to an incorrect conclusion. Tell me more about this project. If anything all you’re really going to learn is that using the wrong data leads to poor results. I’m not sure you’re really going to learn anything about stats going this route.
[/quote]
Well I think he’s only 16, so it’s just for like a high school thing I assume. The results are probably irrelevant so long as he has some kind of data and correctly applies the methods.
[quote]csulli wrote:
The results are probably irrelevant so long as he has some kind of data and correctly applies the methods.[/quote]
These are my thoughts.
Good luck getting a sizeable population of competition lifts where everyone weighs statistically the same, has the same limb proportions, etc. so that you can assess JUST height.
Using just height and looking for a correlation is quite alright for this I would say.
I just found out I’m not allowed to collect my own data due to some bullshit reason. Thanks for all the help guys. If I ever have to collect my own data I’ll know where to go next time. Thanks again
[quote]csulli wrote:
64 424
66 385
66 400
67 490
67 551
68 315
68 450
68 495
69 330
69 375
69 500
69 525
70 150
70 210
70 353
70 410
70 480
70 480
70 490
70 500
70 543
70 555
71 462
71 475
71 584
71 625
72 243
72 410
72 487
72 545
73 440
74 418
74 600
74 625
74 815
75 474
75 510
77 380
77 601
78 455
I did this out of curiosity. Here’s 40 people’s heights(in.) and deadlift maxes(lbs.) from TNation’s training logs. [/quote]
You do that with python?
I prefer SAS myself. It’s a bit more involved, but you have more capabilities and nuancing at your disposal.
Hello, I really like that you are including the topic of powerlifting in your school work.
I was taking an english class a while back and I tired to make most of my essays and paragraphs somehow powerlifting related. Teach probably though I was obsessed :S.
Anyway, I am 5’0 and can deadlift around 250lbs. I have long monkey arms and my legs are half the size of most peoples. HAHA that makes me sound like I look somewhat deformed…
I think this means my deadlift should be higher, but my legs are weak I think.
I know this doesn’t help your project at all, I just really wanted to be a part of this thread.
Thank you and take care.
EDIT: Nevermind didn’t notice OP’s most recent post.