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 10, page257.

Date: March 21, 2004

Question:
 For Exercise 8(b) and (c), describe the control variables that were used to eliminate plausible rival hypotheses.  For each study, what other variables should have been controlled and how would you have controlled them?

Answer 8(b):

In 8(b) control variables about memory recall were embedded in the instructions to those assigned to either of two groups: the group who paraphrased; and the group who used keywords.  Such instructions included (1) telling all participants to read a passage one time and one time only; (2) waiting exactly one week to test recall; and (3) providing everyone 40 minutes to recall the article from memory.

These instructions eliminate other plausible hypotheses.  For example, memory might have been better or worse:
(a) if everyone had been given a time limit within which to read the article as often as possible;
(b) if everyone had read the article aloud;
(c) if everyone had heard the article read, and then studied the article in private; and
(d) if everyone had been told to make a mental flashcard about each key idea while reading and to remember it.

(A) One could control for a time limit by projecting the article on a screen and turning it off after time expired.
(B) Persons could have operated in pairs of two, with one doing the reading, another doing the listening, and by randomly assigning each member of the pair to a corresponding article comparable in written details.
(C) the third hypothesis simply requires the same listening period and study period to enforce.
(D) the final alternative hypothesis suggests a "creative technique" and may or may not appeal to certain participants and therefore produce an effect on memory after a specified period of time.

Answer 8(c):

In 8(c) an experiment was conducted to determine if the educational level of the counselor affects the results of counseling.  This was controlled by (1) having everybody listen to the same tape, even though participants thought the instructor had either more education and training, or less.  By holding everything constant except the variable - teacher experience - the participants' response could be measured in relation to their singular perception about educational level of the instructor.

Counseling effectiveness could be related to many other variables.  To name a few:
(a) mood and spontaneity of the participant;
(b) pleasantness and professionalism of the office surroundings;
(c) self-esteem of the person being counseled.

(A)  One might attempt to control for mood and spontaneity of the participant by serving them a 10-minute cup of coffee and related magazine article, with the same soothing music in the same pleasant surroundings before the interview.
(B)  Again have participants experience the same initial and actual surroundings might influence and therefore be controlled, so as to measure the effect of the counselling on factors other than pleasantness and professionalism;
(C)  Participants could be screened by administering a self-esteem test before the counseling sessions. Persons in roughly the same measurable ranges of high and low self-esteem could be identified in advance and therefore controlled.

filename: StatHW3Ch10Prob10page257TomMeyer.doc

Tom Meyer

Thomas Meyer