Re: spss help

[Art Kendall-2]

I am still unclear what you have. By joints I thought you meant anatomical joints (wrist, elbow, shoulder, hip knee, ankle). Is this it? If not please explain what the 20 joints are.

Do you have 120 separate measures of velocity?

Art

Vikki Mitchell wrote:

Hiya I just have the one subject who did the tennsi serve 20 times, And from the results I have grouped the data into the joints. and i have the impact point for each joint for each trial. so thevelocity of the ball at impact for each of the trials, the velocity of the racquet at impact for 20 trials etc etc and i want to look at the variability of these. so is a one way repeated measures anova correct? and i would do a separate one for each joint?
thankyou for responding

Vikki
> Date: Tue, 25 Aug 2009 09:00:10 -0400
> From: [hidden email]
> To: [hidden email]
> Subject: Re: spss help
>
> If velocity is measured the same way at all 120 times on each subject,
> it appears that you have a two-way anova with 2 repeated measure (within
> subject) variables namely joint and trial. See the GLM repeated
> measures under <help>.
>
> Is you design more extensive than that? How many subjects to you have?
> Are they in any subgroups (between subjects)? treated vs not treated?
> Male vs female? Age 5 vs age 50?
>
> Art Kendall
> Social Research Consultants
>
> ikkivhm wrote:
> > Hello please can someone help me. I am doing my dissertation and am
> > really struggling with my spss test.
> > basically i am looking at the variability of the velocity of joints
> > about the body. i have one subject performing 20 trials. and the
> > output data is velocity of 6 different joints. i was going to do a one
> > way repeated measures anova as my ind variable is trial number and the
> > dependent variables are the 6 sets of velocities. But how could i get
> > all of this into one test, and also would teh 6 different sets of
> > rsults count as levels to the test or not?
> > Please help me
> >

---

Internet Explorer 8 - accelerate your Hotmail. Download Internet Explorer 8

===================== To manage your subscription to SPSSX-L, send a message to [hidden email] (not to SPSSX-L), with no body text except the command. To leave the list, send the command SIGNOFF SPSSX-L For a list of commands to manage subscriptions, send the command INFO REFCARD

[Art Kendall-2]

It has been over 30 years since I dealt with single subject observational designs.  So other list members may want to suggest how to analyze a single subject with 2 within subject factors.  Maybe as 6 separate time-series?


Can you get more subjects so that error terms can be calculated?

There is some syntax below the sig block.
Save all your current work, then open a new instance of SPSS. Make sure that you put warnings, etc. into the output file. <edit> <options> <viewer>. Open your input file. Cut-and-paste then run the syntax.

Note that when you treat that data as "correlated in inclusion" (repeated or within subject) error terms are not present.

If you were to treat the data as if joint were a between subject factor, the error term would not be taking into account the repeated nature of the data.

Also, try running the syntax with different seeds.   Look at the plots.  Notice how even purely random data can appear to have patterns.

Take a big dose of salt with any conclusions. 

Art Kendall


* make up some data.
new file.
set seed = 20090825.
input program.
vector x (120,f3).
loop #p = 1 to 120.
compute x(#p) = rnd(rv.normal(50,10)).
end loop.
end case.
end file.
end input program.
rename vars (x1 to x120 =
 j1t1 j1t2 j1t3 j1t4 j1t5 j1t6 j1t7 j1t8 j1t9 j1t10 j1t11 j1t12 j1t13 j1t14 j1t15 j1t16 j1t17 j1t18 j1t19 j1t20
 j2t1 j2t2 j2t3 j2t4 j2t5 j2t6 j2t7 j2t8 j2t9 j2t10 j2t11 j2t12 j2t13 j2t14 j2t15 j2t16 j2t17 j2t18 j2t19 j2t20
 j3t1 j3t2 j3t3 j3t4 j3t5 j3t6 j3t7 j3t8 j3t9 j3t10 j3t11 j3t12 j3t13 j3t14 j3t15 j3t16 j3t17 j3t18 j3t19 j3t20
 j4t1 j4t2 j4t3 j4t4 j4t5 j4t6 j4t7 j4t8 j4t9 j4t10 j4t11 j4t12 j4t13 j4t14 j4t15 j4t16 j4t17 j4t18 j4t19 j4t20
 j5t1 j5t2 j5t3 j5t4 j5t5 j5t6 j5t7 j5t8 j5t9 j5t10 j5t11 j5t12 j5t13 j5t14 j5t15 j5t16 j5t17 j5t18 j5t19 j5t20
 j6t1 j6t2 j6t3 j6t4 j6t5 j6t6 j6t7 j6t8 j6t9 j6t10 j6t11 j6t12 j6t13 j6t14 j6t15 j6t16 j6t17 j6t18 j6t19 j6t20).



GLM j1t1 j1t2 j1t3 j1t4 j1t5 j1t6 j1t7 j1t8 j1t9 j1t10 j1t11 j1t12 j1t13 j1t14 j1t15 j1t16 j1t17
    j1t18 j1t19 j1t20 j2t1 j2t2 j2t3 j2t4 j2t5 j2t6 j2t7 j2t8 j2t9 j2t10 j2t11 j2t12 j2t13 j2t14 j2t15
    j2t16 j2t17 j2t18 j2t19 j2t20 j3t1 j3t2 j3t3 j3t4 j3t5 j3t6 j3t7 j3t8 j3t9 j3t10 j3t11 j3t12 j3t13
    j3t14 j3t15 j3t16 j3t17 j3t18 j3t19 j3t20 j4t1 j4t2 j4t3 j4t4 j4t5 j4t6 j4t7 j4t8 j4t9 j4t10 j4t11
    j4t12 j4t13 j4t14 j4t15 j4t16 j4t17 j4t18 j4t19 j4t20 j5t1 j5t2 j5t3 j5t4 j5t5 j5t6 j5t7 j5t8 j5t9
    j5t10 j5t11 j5t12 j5t13 j5t14 j5t15 j5t16 j5t17 j5t18 j5t19 j5t20 j6t1 j6t2 j6t3 j6t4 j6t5 j6t6
    j6t7 j6t8 j6t9 j6t10 j6t11 j6t12 j6t13 j6t14 j6t15 j6t16 j6t17 j6t18 j6t19 j6t20
  /WSFACTOR=joint 6 Polynomial serve 20 Difference
  /METHOD=SSTYPE(3)
  /PLOT=PROFILE(serve*joint joint serve)
  /EMMEANS=TABLES(joint) COMPARE ADJ(LSD)
  /EMMEANS=TABLES(serve) COMPARE ADJ(LSD)
  /CRITERIA=ALPHA(.05)
  /WSDESIGN=joint serve joint*serve.


Vikki Mitchell wrote:

Hiya

its hard to explain. but basically i have data from 20 tennis serves by one person. i have looked at the data and calculated the velocity of the specific joint (either ball, elbow, wrist, shoulder, racquet) at the point of impact with the ball. And i want to look at the variability of these results generated. so ihave 20 sets of data for ball velocity at impact, 20 for racquet velocity at impact, 20 for elbow velocity at impact etc etc. Im getting myself all confused lol. But tried to do an anova but dont think it liked it that i only had the one subject's set of data as teh tables didnt look right in the output!
hope this explains it a bit better?
thanks
vikki

---

Date: Tue, 25 Aug 2009 10:36:55 -0400
From: [hidden email]
To: [hidden email]; [hidden email]
Subject: Re: spss help

I am still unclear what you have.  By joints I thought you meant anatomical joints (wrist, elbow, shoulder, hip knee, ankle). Is this it? If not please explain what the 20 joints are.

Do you have 120 separate measures of velocity?

Art

Vikki Mitchell wrote:

Hiya I just have the one subject who did the tennsi serve 20 times, And from the results I have grouped the data into the joints. and i have the impact point for each joint for each trial. so thevelocity of the ball at impact for each of the trials, the velocity of the racquet at impact for 20 trials etc etc and i want to look at the variability of these. so is a one way repeated measures anova correct? and i would do a separate one for each joint?
thankyou for responding

Vikki
> Date: Tue, 25 Aug 2009 09:00:10 -0400
> From: [hidden email]
> To: [hidden email]
> Subject: Re: spss help
>
> If velocity is measured the same way at all 120 times on each subject,
> it appears that you have a two-way anova with 2 repeated measure (within
> subject) variables namely joint and trial. See the GLM repeated
> measures under <help>.
>
> Is you design more extensive than that? How many subjects to you have?
> Are they in any subgroups (between subjects)? treated vs not treated?
> Male vs female? Age 5 vs age 50?
>
> Art Kendall
> Social Research Consultants
>
> ikkivhm wrote:
> > Hello please can someone help me. I am doing my dissertation and am
> > really struggling with my spss test.
> > basically i am looking at the variability of the velocity of joints
> > about the body. i have one subject performing 20 trials. and the
> > output data is velocity of 6 different joints. i was going to do a one
> > way repeated measures anova as my ind variable is trial number and the
> > dependent variables are the 6 sets of velocities. But how could i get
> > all of this into one test, and also would teh 6 different sets of
> > rsults count as levels to the test or not?
> > Please help me
> >

---

Internet Explorer 8 - accelerate your Hotmail. Download Internet Explorer 8

---

Windows Live Messenger: Celebrate 10 amazing years with free winks and emoticons. Get Them Now

===================== To manage your subscription to SPSSX-L, send a message to [hidden email] (not to SPSSX-L), with no body text except the command. To leave the list, send the command SIGNOFF SPSSX-L For a list of commands to manage subscriptions, send the command INFO REFCARD

[Hector Maletta]

This is not a “single subject design”. The unit of analysis is not the person but each tennis serve. There are 20 of them (pretty few, by the way, for a useful statistical analysis). The fact that all come from the same person is irrelevant. There are 20 units of analysis, with a number of variables observed for each unit (i.e. for each serve). That’s enough for a statistical analysis such as correlation, regression or ANOVA (all variants of the same underlying general linear model). The trouble is, of course, that 20 is a small number, and statistics is based on the Law of Large Numbers. Small numbers produce large statistical errors. Get your tennis player to do more and more serves, in order to have a more representative sample of the “population” formed by all its serves.

Hector

 

---

From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of Art Kendall
Sent: 25 August 2009 12:58
To: [hidden email]
Subject: Re: spss help

 

It has been over 30 years since I dealt with single subject observational designs.  So other list members may want to suggest how to analyze a single subject with 2 within subject factors.  Maybe as 6 separate time-series?


Can you get more subjects so that error terms can be calculated?

There is some syntax below the sig block.
Save all your current work, then open a new instance of SPSS. Make sure that you put warnings, etc. into the output file. <edit> <options> <viewer>. Open your input file. Cut-and-paste then run the syntax.

Note that when you treat that data as "correlated in inclusion" (repeated or within subject) error terms are not present.

If you were to treat the data as if joint were a between subject factor, the error term would not be taking into account the repeated nature of the data.

Also, try running the syntax with different seeds.   Look at the plots.  Notice how even purely random data can appear to have patterns.

Take a big dose of salt with any conclusions. 

Art Kendall


* make up some data.
new file.
set seed = 20090825.
input program.
vector x (120,f3).
loop #p = 1 to 120.
compute x(#p) = rnd(rv.normal(50,10)).
end loop.
end case.
end file.
end input program.
rename vars (x1 to x120 =
 j1t1 j1t2 j1t3 j1t4 j1t5 j1t6 j1t7 j1t8 j1t9 j1t10 j1t11 j1t12 j1t13 j1t14 j1t15 j1t16 j1t17 j1t18 j1t19 j1t20
 j2t1 j2t2 j2t3 j2t4 j2t5 j2t6 j2t7 j2t8 j2t9 j2t10 j2t11 j2t12 j2t13 j2t14 j2t15 j2t16 j2t17 j2t18 j2t19 j2t20
 j3t1 j3t2 j3t3 j3t4 j3t5 j3t6 j3t7 j3t8 j3t9 j3t10 j3t11 j3t12 j3t13 j3t14 j3t15 j3t16 j3t17 j3t18 j3t19 j3t20
 j4t1 j4t2 j4t3 j4t4 j4t5 j4t6 j4t7 j4t8 j4t9 j4t10 j4t11 j4t12 j4t13 j4t14 j4t15 j4t16 j4t17 j4t18 j4t19 j4t20
 j5t1 j5t2 j5t3 j5t4 j5t5 j5t6 j5t7 j5t8 j5t9 j5t10 j5t11 j5t12 j5t13 j5t14 j5t15 j5t16 j5t17 j5t18 j5t19 j5t20
 j6t1 j6t2 j6t3 j6t4 j6t5 j6t6 j6t7 j6t8 j6t9 j6t10 j6t11 j6t12 j6t13 j6t14 j6t15 j6t16 j6t17 j6t18 j6t19 j6t20).



GLM j1t1 j1t2 j1t3 j1t4 j1t5 j1t6 j1t7 j1t8 j1t9 j1t10 j1t11 j1t12 j1t13 j1t14 j1t15 j1t16 j1t17
    j1t18 j1t19 j1t20 j2t1 j2t2 j2t3 j2t4 j2t5 j2t6 j2t7 j2t8 j2t9 j2t10 j2t11 j2t12 j2t13 j2t14 j2t15
    j2t16 j2t17 j2t18 j2t19 j2t20 j3t1 j3t2 j3t3 j3t4 j3t5 j3t6 j3t7 j3t8 j3t9 j3t10 j3t11 j3t12 j3t13
    j3t14 j3t15 j3t16 j3t17 j3t18 j3t19 j3t20 j4t1 j4t2 j4t3 j4t4 j4t5 j4t6 j4t7 j4t8 j4t9 j4t10 j4t11
    j4t12 j4t13 j4t14 j4t15 j4t16 j4t17 j4t18 j4t19 j4t20 j5t1 j5t2 j5t3 j5t4 j5t5 j5t6 j5t7 j5t8 j5t9
    j5t10 j5t11 j5t12 j5t13 j5t14 j5t15 j5t16 j5t17 j5t18 j5t19 j5t20 j6t1 j6t2 j6t3 j6t4 j6t5 j6t6
    j6t7 j6t8 j6t9 j6t10 j6t11 j6t12 j6t13 j6t14 j6t15 j6t16 j6t17 j6t18 j6t19 j6t20
  /WSFACTOR=joint 6 Polynomial serve 20 Difference
  /METHOD=SSTYPE(3)
  /PLOT=PROFILE(serve*joint joint serve)
  /EMMEANS=TABLES(joint) COMPARE ADJ(LSD)
  /EMMEANS=TABLES(serve) COMPARE ADJ(LSD)
  /CRITERIA=ALPHA(.05)
  /WSDESIGN=joint serve joint*serve.


Vikki Mitchell wrote:

Hiya

its hard to explain. but basically i have data from 20 tennis serves by one person. i have looked at the data and calculated the velocity of the specific joint (either ball, elbow, wrist, shoulder, racquet) at the point of impact with the ball. And i want to look at the variability of these results generated. so ihave 20 sets of data for ball velocity at impact, 20 for racquet velocity at impact, 20 for elbow velocity at impact etc etc. Im getting myself all confused lol. But tried to do an anova but dont think it liked it that i only had the one subject's set of data as teh tables didnt look right in the output!
hope this explains it a bit better?
thanks
vikki

---

Date: Tue, 25 Aug 2009 10:36:55 -0400
From: [hidden email]
To: [hidden email]; [hidden email]
Subject: Re: spss help

I am still unclear what you have.  By joints I thought you meant anatomical joints (wrist, elbow, shoulder, hip knee, ankle). Is this it? If not please explain what the 20 joints are.

Do you have 120 separate measures of velocity?

Art

Vikki Mitchell wrote:

Hiya I just have the one subject who did the tennsi serve 20 times, And from the results I have grouped the data into the joints. and i have the impact point for each joint for each trial. so thevelocity of the ball at impact for each of the trials, the velocity of the racquet at impact for 20 trials etc etc and i want to look at the variability of these. so is a one way repeated measures anova correct? and i would do a separate one for each joint?
thankyou for responding

Vikki
> Date: Tue, 25 Aug 2009 09:00:10 -0400
> From: [hidden email]
> To: [hidden email]
> Subject: Re: spss help
>
> If velocity is measured the same way at all 120 times on each subject,
> it appears that you have a two-way anova with 2 repeated measure (within
> subject) variables namely joint and trial. See the GLM repeated
> measures under <help>.
>
> Is you design more extensive than that? How many subjects to you have?
> Are they in any subgroups (between subjects)? treated vs not treated?
> Male vs female? Age 5 vs age 50?
>
> Art Kendall
> Social Research Consultants
>
> ikkivhm wrote:
> > Hello please can someone help me. I am doing my dissertation and am
> > really struggling with my spss test.
> > basically i am looking at the variability of the velocity of joints
> > about the body. i have one subject performing 20 trials. and the
> > output data is velocity of 6 different joints. i was going to do a one
> > way repeated measures anova as my ind variable is trial number and the
> > dependent variables are the 6 sets of velocities. But how could i get
> > all of this into one test, and also would teh 6 different sets of
> > rsults count as levels to the test or not?
> > Please help me
> >

---

Internet Explorer 8 - accelerate your Hotmail. Download Internet Explorer 8

 

---

Windows Live Messenger: Celebrate 10 amazing years with free winks and emoticons. Get Them Now

===================== To manage your subscription to SPSSX-L, send a message to [hidden email] (not to SPSSX-L), with no body text except the command. To leave the list, send the command SIGNOFF SPSSX-L For a list of commands to manage subscriptions, send the command INFO REFCARD

No virus found in this incoming message.
Checked by AVG - www.avg.com
Version: 8.5.409 / Virus Database: 270.13.66/2325 - Release Date: 08/25/09 06:08:00