|
|
| Ron Jones Bio |
| CorporateWellness |
| Coach & Train |
| Exercise Library |
| Handouts |
| Health & Fitness |
| KETTLEBELLS |
| Products by RJ |
| Site Map |
|
|
|
|
|
|
-95.jpg){kind=small}
.jpg){kind=small}
Practice
Exam #2/Part A Tip Sheet
ä RULES: (According to my stats tutor, there are certain limits or rules that apply to some of these tests i.e. some will always have more than one IV or multiple levels of the IV, etc. Using these "rules" can help you narrow down which test to use and which ones to disregard for your particular study question)
A.
To test for a difference in the energy cost of exercising on different
types of equipment, the researchers measured oxygen consumption on 24 men
and women who exercised on each of four different types of equipment.
·
DV=V02 (what is going to be
effected by the IVs of equipment & gender)
·
IV#1=Equipment (four levels) and
IV#2=Gender (two levels)
o
This is a 4X2
o
Note that IVs are “factors”
Ø STATS Test #1=Factorial ANOVA (multiple factors involved or “IVs”)
Ø STATS Test #2=Tukey Post Hoc
Ø *If no significant interaction, then look at “main effects”
ä RULE: (1 DV, ³2 IV)
B.
In the same study described above (A), the researchers decided to examine
the effect of gender on the energy cost of exercising on the
NordicTrack cross-country ski simulation machine.
·
DV=VO2 (depends on gender)
·
IV=Gender (two levels with two
means)
Ø STATS Test =Unpaired T-Test (unpaired compares two different uncorrelated populations)
ä RULE: (1 DV, 1 IV with 2 levels)
C.
One technician wanted to know if the rating of perceived exertion
was related to the heart rate during the exercise on the treadmill and if
RPE could be used to predict
heart rate.
·
DV=Heart Rate (depends on IV or
RPE)
·
IV=RPE
Ø STATS Test =Regression Analysis (“Predicts” one variable from another)
Ø
TIPS: If the statement says
“predict” then look at this test and Multiple Regression.
Make sure correlation is significant before you use this test because if
they are not then can’t predict one from the other.
ä RULE: (1 DV, 1 IV with zero levels)
D.
In a study on the differences in one-mile walk times between 5th
grade students in three schools in LA County, the researchers also wanted
to account for possible sex differences that might exist in the walk
time.
·
DV=Walk Times (these depend upon
the IVs of school & gender)
·
IV#1=Schools (three levels) and
IV#2=Gender (two levels)
o
This is a 3X2
Ø STATS Test #1=Factorial ANOVA (multiple factors of school & gender)
Ø STATS Test #2=Tukey Post Hoc
Ø STATS Test #3=If no significant interaction, then look at “main effects”
ä RULE: (1 DV, ³2 IV)
E.
In a diet vs. exercise vs. diet plus exercise study, the three groups
had different mean body mass at the beginning of the study, even though
the participants were randomly assigned. The
researchers will be testing the change in body mass from before to after
a 10-week program involving one
of the three conditions.
·
DV=Body Mass (depends on IV of
10-week program)
·
IV=10-Week Program (two levels of
pre and post test)
Ø STATS Test =Paired T-Test (only testing “one” of three groups)
o You can use One-Way ANOVA but the T-Test is going to be stronger for "rejecting" the null hypothesis i.e. the T-Test is going to be the better choice.
Ø TIPS: You are only testing “ONE” of three groups.
ä RULE: (1 DV, 1 IV with 2 levels)
F.
Andrea is an epidemiologist who is trying to predict the risk
of diabetes in the people of developing nations.
She is measuring body dimensions (height, weight, waist to hip
ratio), physical activity level, dietary levels of whole grains
and age in order to develop a method for estimating individual risk.
·
DV=Risk of Diabetes
·
IV#1=Height, IV#2=Weight,
IV#3=Waist to Hip Ratio, IV#4=Physical Activity Level, IV#5=Dietary Levels of
Whole Grains, IV#6=Age
Ø STATS Test =Multiple Regression (predicts Y from several factors)
ä RULE: (1 DV, >1 IV)
Ron Jones/www.ronjones.org (11-4-01)
Be Strong.