To: sbaker@odu.edu
From: Thomas Meyer
tmeyer@ph.vccs.edu
276 656-0283
Patrick Henry Community College
Subject: Statistics - w/ Dr.Spencer Baker - Homework Assignment #3, Ch 10 - Problem 12, page257.
Date: March 22, 2004
Question:
From the research descriptions provided in Exercises 2-6, 8(b), and 8(c),
indicate which research questions can be determined to have a causal answer, and
justify your answer. Also keep in mind that some of the questions address
more than one research question.
Answer:
To answer exercises 2-6, and 8(b) and 8(c) we can draw upon two paragraphs
on page 254 of the text:
In the first of these paragraphs, its says that:
"The major useful distinction for whether causal inferences can be drawn is
between variables that can be manipulated, or assigned, by the investigator and
variables that are status characteristics of the individual participants.
Manipulated variables are those variable whose values are chosen by the
investigator, and their values can be changed by the investigator. If,
under appropriately controlled conditions, a change in variable X form value X1
to value X2 is consistently followed by a change in variable Y from Y1 to Y2, we
can infer that there is a causal relationship between the change in X and the
change in Y."
But in the second paragraph Thorndike and Dinnel specify that:
"It is seldom if ever possible to conclude that a relationship between status
variables... "(height, weight, gender, intellectual ability, cultural
background, motivation - variables which the participants bring but over which
the investigator has no control) " is causal because the variables cannot be
isolated from their context."
Exercise 2:
Since there are control variables, causation may be inferred. The
expertise of the counselor based on the counseling approach respondent is
manipulated by the researcher. So dependent variables may be expected to
vary in accordance with change in the magnitude of the independent variable.
A change in counseling approach may indeed have a causal result that can be
discovered in relation to the participants' gender. However, the
classification by male and female is a status variable only.
Exercise 3:
Age and gender are status variables, and therefore represent characteristics brought to the experiment by the participants. Since they are not assigned by the researcher, they are basically uncontrolled, and causal inferences are not appropriate.
Exercise 4:
Two strategies for inducing volunteers to donate blood are being tested.
The independent variables include presenting volunteers with factual
presentation only, as well as presenting volunteers with factual presentation
along with visual supplementary information. Since these strategies are
within the power of the researcher to control, there will be definite
possibilities for inferring a causal relationship.
Exercise 5:
This exercise is quite similar to the previous exercise.
Two strategies for improving anger management are being tested. The
independent variables include presenting some participants with rational-emotive
therapy, and presenting participants with rational-emotive therapy coupled with
role playing. The researcher maintains control over which participants are
assigned to either of these strategies and so causal inferences can be made.
Exercise 6:
Unfortunately for the investigator, national origin of culture and gender of the
participants are both status variables. These status variables are brought
to the experiment by the participants and not assigned; therefore no causal
inferences should be made.
Problem 8 (b):
Paraphrasing and the use of key words are two strategies being used to determine
if memory of what has been read can be influenced by strategy. Assigning
participants to either of these strategies was controlled by the researcher.
Therefore causal inferences about the effectiveness of the two strategies on
memory recall are warranted.
Problem 8 (c):
Whether or not the instructor possessed a master's degree, or a doctoral degree
was successfully within the control of the researcher in Problem 8 (c).
Because the presentation was truly done by the same person, and because both
groups of participants were subject to similar controlled time limits, we can
permit the investigator to draw causal inferences.
filename: StatHW3Ch10Prob12page257TomMeyer.doc
Tom Meyer
Thomas Meyer