john2072
czołg
Dołączył: 26 Sty 2011
Posty: 1715
Przeczytał: 0 tematów
Ostrzeżeń: 0/10
Skąd: England
|
 Wysłany: Śro 10:03, 09 Mar 2011 Temat postu: measure of variability = square root of [ |
|
|
y“measure of variability” which is called the
stan-
dard error of the difference in two means.In other words:
measure of variability = square root of [(SEM1)2+ (SEM2)2]
4.A 95% con?dence interval for the difference in the two population means is
difference in sample means ±2 ×measure of variability
or
difference in sample means ±2 ×square root of [(SEM1)2+(SEM2)2]
This format was based on the fact that the “sample value” over repeated samples was
predicted to follow a bell-shaped curve centered on the “population value.” All we
needed to know in addition was the “standard deviation” for that speci?c bell-shaped
curve.
The same is true for calculating the difference in two means. Here is the recipe
you follow:
A DIRECT COMPARISON OF DIET AND EXERCISE
We are now in a position to compute a 95% con?dence interval for the difference
in population means for weight loss from dieting only and weight loss from exer-
cising only. Let’s follow the steps outlinedusing the data from the previous section:
Steps 1 and 2pute sample meansstandard deviationsand SEMs:
Diet OnlyExercise Only
sample mean = 7.2 kgsample mean = 4.0 kg
sample standard deviation = 3.7 kgsample standard deviation = 3.9 kg
number of participan [link widoczny dla zalogowanych]yilai:
[link widoczny dla zalogowanych]
[link widoczny dla zalogowanych]
[link widoczny dla zalogowanych]
[link widoczny dla zalogowanych]
Post został pochwalony 0 razy
|
|
