KINE2049 Research methods
Chapter 5: Experimental research
1. Explain the difference between the terms parameter and statistic
a. Parameters are the facts gathered about the whole population (μ)
b. Statistics are gathered from sample groups (parts of populations) and are regarded as
estimates of parameters (x)
2. Discuss the qualities of proper sampling
a. Clearly identify the population of interest, describing who is and is not included in the
experiment
b. Identify the sample group to avoid any misrepresentation of population (biased)
c. Determine the sample size. Generally, large samples increase the accuracy. However
one must consider the cost, and whether the large sample truly represents the
population. (Literary digest article on presidential survey; large sample however it only
reflected 1 class of people thus it was biased)
d. Randomly select sample where each member has equal chance of selection, and
selection of one does not hinder the probability of others
3. Describe the differences between a random sample, a stratified sample, and a cluster
sample
a. Random sample (simple) – the selection of sample is random and selection of one does
not alter the selection of another
b. Random sample (stratified) – the selection is random, but the final sample would
realistically represent ratios of real populations. Utilized when the ratio is relevant to
study.
c. Cluster sample – selecting samples with unknown characteristics, and when selecting
whole samples are impossible.
4. Explain the difference between an independent variable, a dependent variable, and a
control variable
a. Independent variable – variables manipulated by the investigator to show possible cause
and effect
b. Dependent variable – the variable or outcome measured.
c. Control variable – the variable(s) that must remain constant in order to isolate the
dependent variable
5. Describe why independent variables have levels
Having many levels will allow any independent/dependent correlation to be shown.
6. Explain how operational definitions define variables
a. How the variable (i.e bodyweight) is defined in the research will determine whether it is
independent (weight class of subjects) or dependent (weight gained/loosed by subjects)
7. Understand the threat imposed by confounding variables on research findings
Confounding variables can occur without the knowledge or researchers, and may happen
so discreetly. A valid report must control any possible confounding variables or at least
attempt to identify any variables, failing to do so results in criticism of method. KINE2049 Research methods
8. Differentiate between primary variance, secondary variance, and error variance
a. Primary variance (systematic) – the wanted and consistent variation due in part to the
independent variable. Maximum primary variance can be achieved by having
independent variables change greatly but linearly so correlation can be seen.
b. Secondary variance (extraneous) – the unwanted and consistent variation in the
measurement. Can be controlled with careful selection of subjects, using blind
experimentation, and correctly calibrated instruments.
c. Error variance – unwanted and inconsistent measurement due to uncontrollable
variations or

More
Less