Thursday, October 9, 2014
Chapter 7: Experimental Research
it builds on the principles of the positivist approach more than the others. It is found in a
lot of subjects such as: journalism, criminal justice, education, marketing, nursing, etc.
the basic logic extends from common sense thinking. * view example as it illustrates
three things researchers do in experiments: begin with a hypothesis, modify something
in a situation and compare outcomes with and without the modiﬁcation.
it is the strongest for testing causal relationships because of the three conditions of
causality. (temporal order, association and alternative explanations) are best for
Research Questions Appropriate for an Experiment
-there is no right answer because the question and the technique is not ﬁxed but is
based on informal judgement.
-the questions that are appropriate are those that the logic confronts ethical and
practical limitations of intervening in human affairs for research purposes. The pure
logic has an experimenter intervene induce a change in some focused part of social
life then examine the consequences that result from the change or intervention.
-They cannot manipulate conditions or randomly assign. It is best to be used for
issues that have a narrow scope or scale. this strength allows experimenters to
assemble and run many experiments with limited resources in in a short period.
-better suited for a micro level than for a macro level of concerns for questions. they
can rarely address to look across an entire population. Isolate and target the impact
that arises from the causal variables. Cannot look at dozens at a time because there
-study close related topics using either an experimental or a non experimental
-they frequently want to compare
Thursday, October 9, 2014
-it eases comparison, when doing so they want to compare cases that do not differ
with regard to the variables that offer alternative explanations
Why Randomly Assign?
-it is method to assign cases to groups for the purpose of comparison and it is a
mechanical method as in the assignment is automatic.
-in the experimental design means that every case has an exact exact chance of
ending up in one of the other group. Random Assignment is unbiased because a
researchers desire to conform a hypthesis does not enter in the selection process. It
can be mathematically determined as well. In sampling they select a smaller group
from a big group of a sample.
How to Randomly Assign: begin with a collection of cases and decided into two or more
Matching: it is an alternative to random assignment. It presents a problem because we
don't know the relevance to what extent on characteristics match.
The language of experiments
Parts of the experiment
a) treatment or independent variable - one that gets the modifying
b) dependent variable - physical conditions
c) pretest - the measurement of the dependent variable proper to intro of the treatment
d) post-test - the measurement of the dependent variable after the treatment
e) experimental group - one who receives the treatment
f) control group - does not receive the treatment
g) random assignment
** View box 7.1 on conducting experiments on page 174
Control in Experiments
-they want to control all aspects of the experimental situation to isolate effects on the
treatment and eliminate alternative explanations