false

Unlock Document

Kinesiology & Health Science

KINE 2049

Merv Mosher

Fall

Description

Chapter 6: Experimental Designs
What is an experiment?
In the experiment in the book for the effect of bicycling on weight loss, the
experiment lacks a second level
Dependant variable: body weight, independent variable: exercise program = ,
one-group pre-test-post-test design where a single subject is given a pre-test,
treatment, then a post-test
Problem with this experiment are that there is not a second level of the
independent variable. For a better assessment the program needs a second
level which could be: a control group that does not cycle, or a group that
rides 15 miles a day rather than 10 miles a day. Basically, the original design
lacks a condition against which you can compare the results of the bicycling
program
Must also identify the control variables such as diet, etc. because otherwise
the results may not be because of bicycling alone
One group design:
Bicycling This is the original study. It has
Program Weight 1 independent variable, 1 level,
and 1 dependent variable.
Two group design:
Control Better; it is a 2 level design,
Group Weight subjects can be randomly
assigned to a group and
differences can be because of
the bicycling program.
Bicycling
Program Weight
Advantages of Experiments
Better than just simple observation, more accurate
Experimental control: extraneous variation that may affect the results are
controlled/ eliminated
My identifying many control variables and changing them up in an
experiment, the causal factors can be determined (manipulation of
independent variables, measurement of dependant variables, and control of
other variables)
More than One Independent Variable Advantages: It is more efficient to conduct one experiment with 2-3
independent variables than to conduct many experiments with each single
independent variables (saves time, money, # of subjects)
Experiments with many variables also reflect the real world better and dont
apply just in the lab
Ex. Heart rate is affects by duration of exercise, intensity, and age. Thus, you
must measure all to get an accurate result
Most important reason why multiple independent variables are used is to see
how independent variables interact together and affect the dependant
variable
Disadvantages: as number of independent variables increases, so does the
complexity of the experiment and the interpretation of the results
Complex= 2-3 independent variables, very complex= 4-5
More than One Dependant Variable
Researchers will also measure more than one dependant variable
In bike experiment other dependant variables that can be measures are HR,
lactic acid levels, etc.
Advantages: represents more real life, efficient use of time, subjects,
resources
Disadvantages: increases complexity, more demand placed on subjects, more
interpretation
Experimental Design
Provide the detail and explanation of the different methods used by
researchers to answer question
4 main things it focuses on:
1. Devising subject selection and assignment
2. Writing operational definitions for the independent and dependant
variables
3. Selecting appropriate statistical analysis tests
4. Planning the experimental procedures used to gather the data
Subject Selection and Assignment
3 basic experimental designs:
1. Independent groups design: an experimental design in which each
treatment condition is given to a separate group of subjects. Used when it
is important to exclude subject from the second treatment condition.
Example: 2 groups; one given aspirin other given placebo, thus no one
could take both.
Disadvantage: potential for groups to not be equivalent at beginning of
the study (ex. In aspirin study one must assume that people in both
groups have equal chance of hear attack, etc.)
2. Repeated measures design: the same subjects serve in each condition.
It is an experimental design in which 2 or more treatment conditions are
administered to the same group of subjects. Choosing this design over
the independent groups design depends on the nature of the experiment
and the demands placed on the subjects. Advantage: fewer subjects are needed than with independent groups. It
is also more sensitive than independent groups design to dependent
variable differences. Because the same subjects are in all the treatment
conditions, they are identical in every other aspect. Reduces variablilty
between subjects and improves the chances of detecting any effects
caused by the independent variable a.k.a statistical power
Disadvantages: chance of carry-over effects from one treatment to
another. Carry over effects are present when exposure to previous
treatment affects results for the next treatment.
3. Mixed groups design: combines independent group and repeated
measures methods
Simple Random Assignment
Most common method of assigning subjects to independent variables
Examples: drawing names out of a hat, computer system, etc.
Matched Random Assignment
Has the same goal as simple random assignment, but uses more methods
Used when the researcher wants to make sure that both groups are equal on
the dependant variable
Steps:
1. Select variable that is related to the dependent variable
2. Subjects are then measured on the matching variable
3. Subject with highest scores are paired together and then the 2 groups are
divided
Matched random assignment ensures that groups are equivalent on some
variable prior to the experiment. It doesnt however, make sure that the
groups are equivalent on any other variable
Carry-Over Effects
1. Order Effects: occurs when the result of treatment 1 is still present when
the 2 treatment is given. 2 categories of order effects:
a) Practice effects: results in improved performance on the second of 2
measures because the subject learned how to take the test
b) Fatigue effects: opposite of practice effects, and happens when
performance gets worse on the next tests because of repeated
measurements (sources include: physical fatugue, boredom, loss of
interest)
Counterbalancing
Counterbalancing: controlling for order effects in a repeated measures
design. Counterbalancing the order in which treatments are presented. It
neutralizes the potential order effects in a repeated measures design by
exposing the subject to different order of testing conditions either
systematically or randomly. (pg 106)
With complete counterbalancing, ,all possible orders of presentation are
included in the design of the study. It does become more complex when the
number of treatment conditions goes beyond 3-4.
Latin Squares (pg 107)

More
Less
Related notes for KINE 2049

Join OneClass

Access over 10 million pages of study

documents for 1.3 million courses.

Sign up

Join to view

Continue

Continue
OR

By registering, I agree to the
Terms
and
Privacy Policies

Already have an account?
Log in

Just a few more details

So we can recommend you notes for your school.

Reset Password

Please enter below the email address you registered with and we will send you a link to reset your password.