Test to be conducted

**Mr. Mantis** · Posted On: 10/24/2003 10:52am

The hard and soft styles can't really be separated out can they? I mean, once you start hitting, striking, with power... Soft becomes hard in application right? I suppose it is possible to find people who only trained in tai chi or something similar, it would seem a difficult task to me though. Would your friend be interested in adding a category for people who have a background in both hard and soft work?

I think this is great!

**Ronin** · Posted On: 10/24/2003 10:55am

The soft stylist will be Taiji, Bagua, Xingi.
The theory is that soft stylist hit different than hard, so by comparing we will see if that is so, and if the way they do it is "better", if that is the right word.

**Dochter** · Posted On: 10/24/2003 11:01am

Statistical significane of results with his sample size will be more effected by effect size, the power (1- [Prob of Type II error]), and the variation within treatment groups. With that size of a group there is little chance of a type one error (at least that's how it seems to me).

While any test of this sort is difficult to get generality from (i.e. extend to a population as a whole) 25 individuals in each group is a pretty damn good size for any testing done with large mammals.

Of course you can always work with flies and get ridiculously high sample sizes, or with geese ( I know a guy who is able to study 1/4 of an entire population at a time, how's that for sample size and generality).

**The Wastrel** · Posted On: 10/24/2003 11:13am

Dochter is correct. It's not like you're conducting national opinion polls.

**Ronin** · Posted On: 10/24/2003 11:18am

So the size of the groups seems to be ok?

**The Wastrel** · Posted On: 10/24/2003 11:25am

I think so...But I don't do kinesiological research. It seems to me though that the number of measured strikes should be higher. And you have to account for fatigue right?

**Dochter** · Posted On: 10/24/2003 11:29am

I personally don't have a problem with it and I don't see anything statistically wrong with it (but I'm not a great statistician, or even good).

As I said any "significance" seen will be because there is actually a pretty solid difference between the groups.

With tests like this (moreso with medicine and ecology etc..) the question of significance isn't as clear cut because of what you are looking at. It all boils down to what sort of effect size you are interested in and how big/small you are able to test for. Some people are starting to think that significance testing needs to be completely dumped because of the misinformed reading that just because something fails to be "significant" at a 0.05/5% level it doesn't exist. In actuality it just means that your results could have come about by chance. Simulation modelling has shown that two distinct populations can often give this result due to lack of power in testing procedures. In actuality dropping sig testing is too dramatic and maybe a reanalysis of how we approach problems is instead needed. Additionally, alternative methods of statistical testing are being pursued, though it will probably be awhile till most of us start using them.

**Dochter** · Posted On: 10/24/2003 11:31am

In rereading what I wrote, disregard it no one will care.

25 replicates per blocking factor with repeated measures is a perfectly fine experimental design.

**Ronin** · Posted On: 10/24/2003 11:34am

I think, more than anything, it will be interesting to see what differnces in impact force different striking styles can produce.
As for the size comparison, I am not sure if that will resolve any debate about size.
But to compare Supposedly different ways of striking sounds intriguing.

**Dochter** · Posted On: 10/24/2003 11:44am

Yeah, you won't be able to make any big claims about size but perhaps somethings about power generation. It is a neccesary blocking factor though.

I forget, are the individuals all from different schools and gyms or the same ones. If from the same its a pretty clearcut case of pseudoreplication. In other words all you would be able to say is that the different schools/gyms in particular are different, not anything about hard and soft styles.